Two-Stage Robust Optimal Scheduling of “Nearly-Zero-Carbon Park” Considering the Synergy of Waste Disposal and Carbon Capture Gas-Fired Power Plants

Aiming at the demand characteristics of “Nearly-zero Carbon Parks" (NZCP) powered by all renewable energy for low-carbon emissions and waste reduction, a NZCP topology and a two-stage robust optimal scheduling model based on the Carbon Capture Gas-fired Power Plant (CCGPP)-Power to Gas(P2G)-Waste Disposal (WD) collaborative model are proposed in this paper. First, the relationship between WD, CCGPP and energy supply and demand that can be generated from renewable energy is studied, and according to the system energy regulation requirements, the energy flexibility of residential loads, small industrial loads and electric vehicles are considered, and the NZCP mathematical model based on the CCGPP-P2G-WD collaborative model is established. Secondly, NZCP energy supply and demand characteristics are studied, and an energy coordination model for NZCP energy conversion and storage powered by 100% renewable energy is established. Then, wind power, photovoltaic power, and combined demand response uncertainties are considered, and the NZCP two-stage robust hybrid optimal dispatch model is established based on second-order cone relaxation and dual theory. Finally, the simulation analysis is carried out based on the operation data of the integrated energy system of a park in a certain area in northern my country. The simulation results show that the proposed two-stage robust optimal scheduling model of NZCP coordinated with WD and CCGPP can effectively improve the economy of the system at a higher energy balance level, improve the absorption capacity of renewable energy, and provide better multi-time scale energy regulation characteristics for NZCP with a larger energy supply scale.


INTRODUCTION
In the context of the zero-carbon strategy and the "dual-carbon" goal under the "14th Five-Year Plan" modern energy system plan, seeking low-carbon, diverse, interactive and sustainable energy coupling and transportation is the key to solving the problems of low-carbon development and transformation of energy in human society (National Development and Reform Commission, 2022).The "China Research Report on Carbon Neutrality Before 2060" pointed out that carbon capture, utilization and storage technology and biomass carbon capture and storage technology are important ways and key means to achieve the goal of carbon neutrality. Carbon capture and storage technology can effectively reduce carbon emissions and achieve zero and negative carbon emissions goals Hu et al., 2021;Wang et al., 2021a). However, most of the carbon capture power plants based on the transformation of traditional coal-fired power plants adopt the split-flow operation mode, and the working process of the absorption tower and the regeneration tower is coupled, resulting in the higher the power output, the greater the energy consumption of carbon capture, and the lack of carbon utilization mechanism. Therefore, considering the negative carbon emission characteristics of biomass waste treatment as a zero-carbon renewable energy and the energy space-time translation characteristics of P2G equipment with both source and load attributes, the establishment of NZCP based on CCGPP-P2G-WD collaborative mode is the key to improve the consumption of renewable energy, reduce carbon emissions and realize carbon recycling.
Carbon capture and storage technology is one of the main lowcarbon technologies at the current stage, which can absorb abandoned wind and solar energy and achieve carbon emission reduction. At home and abroad, some researches have been carried out on the flexible operation mode of carbon capture power plants. Cui et al. (2021a) introduces a carbon storage device into the flue gas bypass of a carbon capture coal-fired power plant, shifts the wind power output in time, and proposes a flexible operation mode of coordination between wind power and carbon capture power plants, which improves the wind power consumption capacity; Cui et al. (2021b) introduces price-based demand response into a comprehensive energy system with carbon capture equipment, and adjusts the load to match the output of renewable energy, which reduces the carbon emissions of the system and improves the operating economy of the system;  proposed a combined operation mode of power-to-gas-carbon capture, which uses CO 2 captured by carbon capture equipment as a raw material for methane production in the power-to-gas process, which improves the carbon utilization level on the basis of ensuring economical efficiency.
NZCP fully dispatches zero-carbon resources in the comprehensive energy system of the park, and realizes rapid reduction of carbon emissions and approaches zero through multi-energy coupling and complementation, multi-load demand response, waste treatment and carbon capture and storage technologies. As a typical comprehensive energy application scenario, the park has carried out a lot of researches at home and abroad on the energy balance uncertainty of the park-level comprehensive energy system and the improvement of the energy supply regulation capacity.  established the topological structure of the integrated energy system of the park, considering the economic indicators of environment and energy efficiency, and proposed an optimization method of the integrated energy system of the park based on the improvement of efficiency and benefit; Fang et al. (2020) analyzes the time-delay characteristics of thermal loads, considers the comprehensive demand response mechanism of various electric heating loads, improves the flexibility of energy supply of the comprehensive energy system in the park, reduces wind and light abandonment, and realizes the coordination and optimization of the comprehensive energy system in the park; Jiang et al. (2021) considers the randomness of renewable energy output and load demand in multi-energy parks, and proposes an optimal scheduling method based on an improved deep deterministic policy gradient algorithm, which reduces the impact of source and load uncertainty on scheduling; According to fuzzy theory, Zhong et al. (2020) expresses wind power and load with fuzzy parameters, and transforms deterministic system constraints into fuzzy chance constraints to solve the source and load uncertainty problem in the optimal dispatch of carbon capture virtual power plants.
Biomass energy is also an important zero-carbon energy source, and a lot of efforts have been made at home and abroad in terms of WD energy supply characteristics and optimized operation. Wang et al. (2021b) studied the energy supply characteristics of waste incineration and sewage gas production, considering the processing capacity and waste output of waste treatment facilities, and proposed an operation strategy for waste treatment facilities to participate in power grid peak regulation; Yang et al. (2021) considers the low-carbon characteristics of waste treatment units and introduces a carbon quota trading mechanism, which improves the consumption capacity of renewable energy while taking into account the waste treatment; Zhou et al. (2019) considered the high energy consumption characteristics of flue gas treatment in waste incineration power plants, decoupled the power generation process from the flue gas treatment process, and used abandoned wind for flue gas treatment, which improved the operation flexibility of waste incineration power plants with a high proportion of wind power connected.
From the above analysis, it can be seen that the current domestic and foreign research on the energy balance characteristics and adjustment capacity of the integrated energy system under the low-carbon constraints mainly focus on the carbon capture technology of coal-fired power plants and the low-carbon economic dispatch considering the carbon quota trading mechanism. On the other hand, there are few studies on the collaborative optimization of CCGPP and WD and the improvement of power grid, heat grid, and gas grid regulation capacity.
Therefore, the NZCP structure based on the CCGPP-P2G-WD synergy model is proposed in this paper, which adopts a fully renewable energy supply method and makes full use of zerocarbon resources such as waste incineration, biogas treatment and multi-load demand response in the park. In addition, the energy supply and demand characteristics of NZCP are deeply analyzed. Firstly, the energy production and conversion model of each equipment is established based on the CCGPP-P2G-WD collaborative model. Secondly, an energy coordination model of NZCP energy conversion and storage is established by using the energy coupling matrix to map the multi-energy coupling method. Then, the NZCP two-stage robust optimal scheduling method is proposed, and the two-stage robust optimal scheduling model is processed based on the second-order cone relaxation and dual theory. Finally, the NZCP simulation model is established, and four scenes are proposed for simulation comparison analysis. The validity and economy of the model proposed in this paper are verified.

NEARLY-ZERO CARBON PARKS SYSTEM MODEL
The proposed NZCP considering the synergy of waste treatment and carbon capture gas-fired power plants is aimed at the zerocarbon demand and waste treatment demand of the integrated energy system of the park, and considers the energy transfer characteristics of waste incineration flue gas treatment and carbon capture gas-fired power plants carbon storage, which is sufficient. Coordinate zero-carbon resources such as waste treatment and demand-side response in the park, based on the coordinated coupling system between carbon capture, storage and utilization, energy storage medium material flow and energy flow. The CCGPP-P2G-WD collaborative method with multitemporal scale and multi-energy regulation ability constructed in this paper and the NZCP energy coupling mode and connection relationship it serves are shown in Figure 1. NZCP includes renewable energy supply systems such as wind power generation, photovoltaic power generation, waste incineration and manure treatment, and the entire energy interaction network is composed of power grid, heat network, gas network and transportation network. NZCP realizes the recycling of carbon flow in the energy coupling link through the synergy of CCGPP, WD and P2G.

Waste Treatment System Model
For the waste treatment requirements of NZCP, the classified combustible waste is transported to a waste incineration power plant for waste incineration to generate electricity (Wang et al., 2021c). Biomass technology is used to convert the dry-wet separation of manure and waste into biogas, which is then purified and processed to obtain natural gas, which is used as the input of carbon capture gas-fired power plants and connected to the natural gas network at the same time Teng et al., 2021).
Considering that in the waste treatment system, waste incineration and manure treatment can support the multisource regulation capacity of electric heating gas in the park, the energy conversion model is as follows: where P WD,E , P WD,H and P WD,G are the electric energy, heat energy and natural gas energy provided by the waste treatment system respectively, η WI,E and η WI,H are the power supply and heating efficiency of the waste incineration power plant respectively, a BG,E and a BG,H are the conversion coefficients of electric energy and heat energy of the waste incineration power plant respectively, η FW,G and η SG,G are the treatment efficiency of the excrement treatment facility and the treatment efficiency of the sewage treatment facility respectively, η F,G and η S,G are the conversion coefficient of fecal waste and sewage waste into natural heat and gas energy respectively, a F,G is biogas methanation efficiency, P BG , P FW and P SG are the input amount of combustible waste, fecal waste and sewage waste, respectively. During the working process of the waste incineration power plant, the exhaust gas needs to be discharged after the flue gas treatment reaches the standard (Xv et al., 2018). However, the energy consumption of flue gas treatment accounts for nearly 1/4 of its power generation, and if it is classified as plant power, it will affect the operating economy of the system (Cao et al., 2020). Therefore, a flue gas storage device was introduced to decouple the waste incineration power generation process from the flue gas treatment process. The flue gas storage device divides the flue gas discharged from the waste incineration power plant. By adjusting the proportion of the flue gas entering the reaction tower and the flue gas storage device, the energy demand of the flue gas treatment can be shifted in time and space to realize the dynamic flue gas treatment at different time scales.
The amount of flue gas generated by waste incineration power generation can be expressed as: where v S,WI is the unit flue gas coefficient of waste incineration power generation, V S,RT is the amount of flue gas directly entering the reaction tower, V S,MIGS is the amount of flue gas entering the gas storage device.
In addition to the energy consumption of flue gas treatment, there is energy consumption of the air pump in the filling and deflating behavior of the flue gas and the gas storage tank. The total energy consumption of flue gas treatment in waste incineration power plants can be expressed as: where s S,RT is the energy consumption coefficient of the reaction tower, s S,MIGS is the energy consumption coefficient of the air pump, V S,MIGS−RT is the amount of flue gas entering the reaction tower from the flue gas storage device.

CCGPP-P2G-WD Collaborative Mode
In the CCGPP-P2G-WD cooperative operation mode, the CO2 emitted by CCGPP-WD provides P2G with hydrogen methanation raw materials, and P2G-WD provides CCGPP with natural gas as fuel to realize the electricity-carbonelectricity cycle process. During the working process of carbon capture gas power plants, carbon capture energy consumption is very high. In this paper, the carbon capture gas-fired power plant adopts the solution storage operation mode, and its specific operation mode and electric carbon characteristics refer to reference (Chen et al., 2012). The CCGPP power generation process is decoupled from the CO 2 absorption process by introducing a solution reservoir. The spatiotemporal shift of the energy demand of CCGPP-P2G is achieved, and the rate of CO 2 absorption and desorption is controlled.
The total amount of CO 2 captured by CCGPP-P2G at the moment is: where θ CO 2 ,t is the carbon emission intensity processed at time t, η CCGPP−CO 2 ,t is the carbon capture rate of CCGPP capture CO2 at time t, P GD,E,t is the electricity provided by the waste treatment power plant at time t, P CCGPP,t is the output power of CCGPP at time t. The total amount of CO 2 consumed by P2G at time t is: where χ CO2,t is the amount of CO 2 required to generate unit capacity of natural gas, η P2G is the gas production efficiency of P2G, P C,P2G,t is the energy consumption of the P2G device at time t. The gas production of P2G at time t is: where H g is natural gas calorific value. The total energy consumption of CCGPP-P2G at time t is: where P C,CCGPP,t is the energy consumption of CCGPP at time t, P C,P2G,t is the energy consumption of P2G at time t, P F,t is the energy consumption of system at time t. The net output of CCGPP and net CO 2 emissions at time t are: where P CCGPP,t is the total output of CCGPP at time t, q CO 2 is the amount of CO 2 produced by unit power generation output.

Load Demand Response Model
Load demand response is an important zero-carbon resource, which can optimize the low-carbon performance of the mining system through source-load coordination and has good dispatchability (Wei et al., 2018). The main loads in the park include: smart users, small industrial loads, commercial loads and electric vehicles. Under the limitation of the scope of the park, the energy demand of electric vehicles is high, the battery capacity is small, and the scale of cluster operation is insufficient. This paper does not consider electric vehicles to discharge the grid. Smart users can adjust energy consumption behavior according to changes in electricity prices and reduce energy demand. The energy consumption period of small industrial loads is flexible, and the energy consumption period can be shifted. Commercial loads have a large energy demand. In addition to rigid demand response resources such as lighting, elevators and computers, flexible loads such as air conditioning and energy storage in commercial buildings can be mobilized to participate in the demand response to effectively improve the multiple regulation capacity of the park and reduce the peak-valley difference of load curves (Gao et al., 2019). The types of demand response are divided into price demand response and incentive demand response. Price-based demand response guides users' energy consumption behavior through electricity prices, and the participants are mainly translatable loads (Wang et al., 2021d). Incentive demand response restricts users' energy consumption behavior through contracts, and the participants are mainly interruptible loads (He et al., 2019). Responding to changes in energy prices through load reduction and translation can match the energy demand of the park and promote the level of renewable energy consumption. The NZCP load demand response model is as follows: where i is the load type, that is, the electrical load, the thermal load, and the gas load respectively, k is the area type, residential area, small industrial area, commercial area and electric vehicle, respectively, L k,i,DR,t , L int k,i,DR,t , L shi k,i,DR,t , L k,i,t and L k,i,t, max are the load involved in demand response, interruptible load, transferable load, load forecast value, and maximum allowable load at time t in the k area respectively, α int k,i,o,t and α shi k,i,t are the interruptible load proportion of the oth stage at time t in the k area, and the transferable load proportion at time t in the k area respectively.

NEARLY-ZERO CARBON PARKS ENERGY SUPPLY AND DEMAND CHARACTERISTICS
The proposed NZCP takes the CCGPP-P2G-WD collaborative energy supply system as the core, takes wind power generation and photovoltaic power generation as the main energy supply methods, and realizes the conversion of various energy sources through P2G, CCGPP and Electric Boiler (EB), supplemented by the coordination of electricity storage, heat storage and gas storage devices.
The energy coupling matrix is used to describe the multienergy conversion, storage and distribution relationship of NZCP: where P E,W , P H,W and P G,W are the electricity, heat and gas power purchased from the main network respectively, P ES , P HS and P GS are the electricity, heat and gas power provided by the electricity storage, heat storage and gas storage devices respectively, η GE , η GE,H , η EB and η P2G are CCGPP power supply efficiency, CCGPP thermoelectric ratio, EB heating efficiency, and P2G gas supply efficiency, ς WP,1 and ς WP,2 are the distribution coefficient of wind power supply EB and P2G respectively, ς PV,1 and ς PV,2 are the distribution coefficients of PV supply EB and P2G respectively, P WD,EW , P WD,HW and P WD,GW are the WD power supply, heating power, and gas supply power respectively, P GE a is the total output of CCGPP, P WP,EW and P PV,EW are the on-grid power of wind power and the on-grid power of photovoltaics respectively.

Objective Function
In this paper, the NZCP optimal scheduling model considering the coordination of WD and CCGPP is established. Its essence is to realize carbon recycling through ccGPP-P2G-WD cooperative mode on the basis of considering the waste treatment and carbon reduction requirements of the park, and form a comprehensive energy system with near zero carbon emissions. The optimal goal of minimum net operating cost of the system was established, and the decision variables included total CCGPP output, WD equivalent output, P2G output, EB output, carbon capture energy consumption, flue gas treatment energy consumption, transfer load of demand response, interruption load of demand response and energy purchase from the main network. where C s−s , C buy , C WP,PV , C L,DR , C EB , C P2G , C GD , C CCGPP , C CSE and R CCGPP are the total equipment start-up and shutdown cost, the main grid energy purchase cost, the operation and maintenance cost of renewable energy power generation equipment, the transferable load and interruptible load compensation cost, the EB operating cost, the P2G operating cost, the WD operating cost, the CCGPP operating cost, the carbon sequestration cost , and the income obtained by the carbon capture gas power plants by selling the emission credits that do not meet the carbon emission quotas. ep i,t is the start-stop state of the controllable device ep i at time t, γ ep i ,t is the start-stop cost of the controllable device ep i at time t, z E,t , z H,t and z G,t is the price of electricity, heat and natural gas at time t, γ WP,t and γ PV,t are the unit operation and maintenance costs of wind power and photovoltaics at time t respectively, b k,i,o,int,t and b k,i,shi,t are the electric/heat/gas load interruption compensation price for the oth level participating in demand response at time t in the k area, and the electric/heat/gas load transfer compensation price for participating in the demand response at time t in the k area, γ EB,t and γ P2G,t are the operating cost coefficients of EB and P2G respectively, a, b, c are the cost factor for waste disposal, γ CCGPP,t is the operating cost coefficients of CCGPP, γ CSE,t is the unit cost of carbon sequestration, Q CSE,CO 2 ,t is the amount of CO 2 sequestered by the carbon sequestration device; Q P2G−CO 2 ,t is the amount of CO 2 captured for P2G devices, γ C,t is the carbon trading price, a mount is the carbon emission benchmark quota per unit of electricity.

Restrictions
(1) Power balance constraints: L E,t P E,W,t + P WP,t + P PV,t + P WD,EW,t + P CCGPP,E,t + P ES,t − P P2G,t − P EB,t (28) L H,t P H,W,t + P WD,HW,t + P CCGPP,H,t + P HS,t + P EB,t (29) L G,t P G,W,t + P WD,GW,t + P P2G,H,t + P GS,t − P CCGPP,t where P WP,t and P PV,t are the output power of wind power and photovoltaic generators at time t respectively, L E,t , L H,t and L G,t are the demand for electricity load, heat load, and gas load at the moment respectively, P CCGPP,E,t and P CCGPP,H,t are the power supply and heating power of CCGPP, respectively.
(2) CCGPP operating constraints: where P CCGPP,max and P CCGPP,min are the upper and lower limits of CCGPP output respectively, ΔP CCGPP is the absolute value of the CCGPP ramp rate.
(3) WD operating constraints: WD includes waste incineration power plants and manure treatment facility. The operation constraints of waste incineration power plants are: ΔP BG, min ≤ ΔP BG,e,t − P BG,e,t−1 ≤ ΔP BG, max 0 ≤ P BG,e,t ≤ P BG,e,rated where ΔP BG, max and ΔP BG, min are the upper limit and lower limit of the ramp rate of the waste incineration power plant respectively, P BG,e,t is the output power of the waste incineration power plant, P BG,e,rated is the rated power of the unit of the waste incineration power plant. The operation constraints of manure treatment equipment are: 0 ≤ P FS,g,t ≤ P FS,g,rated where P FS,g,t is the output power of the manure treatment equipment at time t, ΔP FS, max and ΔP FS, min are the ramp rate constraint of the manure treatment equipment, P FS,g,rated is the rated power of the manure treatment equipment.
(4) EB output and climbing constraints: where P EB,max and P EB,min are the upper limit and lower limit of EB output respectively, ΔP EB is the EB ramp rate constraint.
(5) EB output and climbing constraints: where P P2G,max is the upper limit of P2G output, ΔP P2G is the P2G ramp rate constraint.
(6) Energy storage device operating constraints: where P ES,t−1 , P HS,t−1 and P GS,t−1 are the electricity storage, heat storage and gas storage capacity of the electricity storage device, heat storage device and gas storage device at time t-1, respectively, ς ES , ς HS and ς GS are the energy loss rates of electricity storage devices, heat storage devices, and gas storage devices, respectively, η ch ES and η dis ES are the charging efficiency and discharging efficiency of the power storage device, respectively, η ch HS and η dis HS are the charging efficiency and discharging efficiency of the heat storage device, respectively, η ch GS and η dis GS are the charging efficiency and discharging efficiency of the gas storage device, respectively, P ch ES,t and P dis ES,t are the charging power and discharging power of the power storage device, respectively, P ch HS,t and P dis HS,t are the charging power and discharging power of the heat storage device, respectively, P ch GS,t and P dis GS,t are the charging power and discharging power of the gas storage device, respectively.

NEARLY-ZERO CARBON PARKS TWO-STAGE ROBUST OPTIMAL SCHEDULING MODEL Optimized Scheduling Model
NZCP is powered by all renewable energy sources, but wind power, photovoltaic power and multiple loads have strong uncertainties. According to the traditional polyhedral uncertainty set form, the uncertainty factors are expressed as (Sun et al., 2020): where P WP t ,P WP t and ΔP WP t are the output uncertainty, forecast value and fluctuation deviation value of wind power at time t respectively, P PV t ,P PV t and ΔP PV t are the PV output uncertainty, predicted value, and fluctuation deviation value at time t respectively, L i,t ,L i,t and ΔL i,t are the uncertainty, predicted value and fluctuation deviation value of i load at time t respectively, i ∈ [E, H, G], Γ WP , Γ PV and Γ L i are the uncertain adjustment parameters introduced for wind power output, photovoltaic output, and load power, respectively, the value range is an integer within the period 0-24.
After the polyhedron uncertainty set is established, the model changes from a single stage to a two-stage model, and Eq. 16 is transformed into the following two-stage robust hybrid optimal scheduling model: where x is the first-stage optimization variable, corresponding to the start-stop state of each controllable device, the start-up state of the equipment is 1, and the stop state is 0, y and z is the second-stage optimization variable, representing the output, wind, and load uncertain variables corresponding to each equipment respectively, c and b is the coefficient matrix corresponding to the objective function, A, B, C, M, N and Q are a sparse matrix of variables under the corresponding constraints, h, d, u and q are constant column vector. The established NZCP two-stage robust optimal scheduling model decomposes the two-stage problem into the first-stage main problem and the second-stage sub-problem. The main problem is to optimize the start-stop state of each controllable device at each time period under extreme conditions of wind and solar output and original load. The sub-problem is based on the optimization scheme obtained from the main problem, and the output of each equipment in the NZCP is adjusted according to the uncertain variables of wind power, photovoltaic power generation and load to achieve the optimization goal of minimum net cost.

Solution Method
The C&CG algorithm is used to decompose the above two-stage robust optimal scheduling model into a main problem and subproblems (Liu et al., 2018). The main problem is: where k is the current iteration number, y k is the variable at the kth iteration, z k is the value of the uncertain variable z in extreme scenarios obtained after the kth iteration.
The sub-problems is: In a given scenario, the inner min problem is a second-order cone programming problem, which is obtained by combining dual theory with the outer max problem: When Fq. 56 takes the maximum value, the value of the uncertain variable z should be at the boundary of the uncertainty set fluctuation interval established in this paper . At the same time, when the wind and solar output is at the lower limit and the load is at a peak, the system operating cost is the highest, that is, in an extreme scenario, which can be expressed as: where z WP , z PV and z i,L are the binary variable. When the value is 1, it means that the boundary of the interval is obtained. At this time, Eqs 56, 57 are rewritten as: where ϑ is the upper bound of the dual variable. After the above derivation and transformation, the two-stage robust optimization scheduling model is finally decoupled from the main problem (52) and the sub-problem (59), and then solved by the C&CG algorithm, in which the sub-problem can be solved by calling the Cplex solver on the Matlab platform, the algorithm flow is shown in Figure 2. CCGPP carbon emission intensity a; P2G-CCGPP fixed energy consumption b; carbon emission allowance price c.

CASE SIMULATION Example Basic Data
The simulation analysis is based on the operation data of the Park Integrated Energy System in a region of northern China. In NZCP, the forecast curves of wind power and photovoltaic power generation are shown in Figure 3; The daily load forecast curve is shown in Figure 4; the energy price forecast curve is shown in Figure 5; the operating parameters of the energy conversion equipment and energy storage equipment are shown in Table 1; The daily load forecast curve is shown in Figure 4; the energy price forecast curve is shown in Figure 5; the   operating parameters of the energy conversion equipment and energy storage equipment are shown in Table 1; CCGPP carbon emission intensity θ CO 2 ,t 0.76/MW · h; P2G-CCGPP fixed energy consumption P F,t 15MW; carbon emission allowance price γ C,t 129E/t; The waste treatment capacity of Waste Incineration power generation equipment is 1500t/d, and the rated power is 30 MW; The manure treatment capacity of the Manure Treatment facility is 350t/d, and the rated power is 6 MW.

Optimizing Results Analysis
In order to verify the effectiveness of the proposed CCGPP-P2G-WD collaborative model and the NZCP two-stage robust optimal scheduling model, the following four scenes are set up to simulate and analyze the optimal model proposed in this paper: Scene 1. Only P2G is considered to participate in the coordinated operation of the park's comprehensive energy system, and all energy shortages are purchased from the external network.
Scene 2. WD is added On the basis of Scene 1, but the synergy between its energy transfer characteristics and the comprehensive energy system of the park is not considered.
Scene 3. CCGPP is added on the basis of Scene 1, but the synergy between its energy transfer characteristics and the comprehensive energy system of the park is not considered.
Scene 4. Adopt the proposed CCGPP-WD-P2G collaborative operation mode, and use the proposed operation strategy to coordinate and optimize with the park's comprehensive energy system. On the basis of meeting the NZCP load requirements, the multi-energy optimization operation effect of each device in Scene 4 is shown in Figures 6-8.
In Scene 4, the CCGPP-WD-P2G collaborative mode provides certain energy support for the power grid, heat grid, and gas grid. Combined with the forecast curve of electricity, heat and gas load of renewable energy, it can be seen that in the period of 16-18 and 23-6, the system is in a low period of electricity consumption, and the power generation of renewable energy cannot be fully absorbed. Because of the introduction of a flue gas storage device, the high energy consumption of flue gas treatment of WD provides a way to absorb abandoned wind. At the same time, CCGPP transfers the energy consumption of carbon capture during the peak period of energy consumption to the period of low load by introducing solution storage, which further absorbs the abandonment of wind and light on the basis of ensuring the reliability of energy supply during the peak period. Then, the CO 2 captured by CCGPP is used as the raw material in the P2G production process, and the waste electricity is fully absorbed by P2G. It can be seen from Figure 5 that the electricity price at this time is relatively low. On the basis of fully absorbing the abandoned wind and solar energy, the electricity storage equipment is charged by purchasing energy from the external network to adjust the load peak period.
During the period of 7-10, the output of wind power decreases. It can be seen from Figures 6-10 that CCGPP-WD is the main power source at this time. At the same time, the energy consumption of carbon capture and flue gas treatment is transferred to the load trough period, which improves the energy supply efficiency of CCGPP-WD. The heat load demand is low, and the heat energy provided by the CCGPP-WD is stored by the heat storage device. During the 18-6 period, the heat load is at the peak period, and the heat storage device cooperates with CCGPP-WD and EB to supply energy for the heat load. The manure treatment equipment in WD obtains natural gas by purifying biogas, and P2G converts electricity into natural gas through CO 2 captured by CCGPP as a raw material. These two types of gas sources provide natural gas for CCGPP. The gas shortfall is purchased from the mainnet.
Combining Figures 6-8, it can be seen from the optimized load curve that, compared with the load prediction curve before optimization, the load peak-to-valley difference in Scene 4 is smaller. Therefore, the CCGPP-WD-P2G cooperative mode realizes load shaving peaks and valleys.
From Figures 9, 10, it can be seen that during the period 10-16, the output of photovoltaic power generation is mainly used for carbon capture and flue gas treatment. The volatility problem of photovoltaics is well resolved by the CCGPP-WD-P2G synergistic mode introduced into solution storage and flue gas storage devices.
It can be seen from Figure 11 that in Scene 1, P2G and energy storage devices are limited by capacity constraints and ramp rate constraints, resulting in a large amount of power abandonment, and the power abandonment rate reaches 40.19%. In Scene 2, the WD flue gas treatment process consumes a certain amount of abandoned wind and light, and the abandoned electricity rate is 24.19%. In Scene 3, the carbon capture process of CCGPP absorbs a lot of abandoned wind and light, and the electricity abandonment rate is 16.57%. Scene 4 adopts the CCGPP-WD-P2G collaborative mode to shift the energy consumption of carbon capture and flue gas treatment in time and space, and mainly uses abandoned wind and sunlight for energy supply, so the system has the strongest ability to absorb renewable energy in this scene, the power abandonment rate is only 5.33%. It can be seen that the CCGPP-WD-P2G collaborative mode improves the renewable energy consumption capacity on the basis of ensuring the balance of energy supply and demand in the system. The optimization results of each unit cost and total net cost of the four scenes proposed in this paper are shown in Table 2. Compared with Scene 1, the introduction of WD reduces the cost of electricity purchase by 7.96×10 3 ¥, the heating cost of 6.29×10 3 ¥ and the gas purchase cost of 5.79×10 3 ¥, and the total cost is reduced by 3.72×10 3 ¥. Combining with Table 2 and Figure 12, it can be seen that although WD provides energy support for the power grid, heat grid and gas grid in the process of realizing waste reduction, the    carbon emission increases by 64.47%. Compared with Scene 1, in Scene 3, the electricity purchase cost is reduced by 23.03×10 3 ¥, the heat purchase cost is reduced by 10.33×10 3 ¥, and the gas purchase cost is increased by 8.99×10 3 ¥. It can be seen that CCGPP provides a certain energy support for the system, but the gas source in the system is insufficient, which increases the cost of gas purchase. Combined with Table 2 and Figure 12, it can be seen that after the introduction of CCGPP, carbon emissions were reduced by 41.83% through carbon capture and carbon sequestration. Combining Table 2 and Figures 11, 12, it can be seen that Scene 4 adopts the CCGPP-WD-P2G collaborative mode proposed in this paper, which reduces the cost of electricity purchase by 42.78×10 3 ¥, the cost of sufficient heat by 13.49×10 3 ¥ and 2.84×10 3 ¥ gas purchase cost. It can be seen that carbon capture provides P2G with a large amount of CO 2 as a raw material for gas production, which saves gas purchase costs and P2G operating costs. Because carbon capture and carbon sequestration technology reduces carbon emissions, NZCP's participation in carbon emission allowance trading increases revenue by 9.724×10 3 ¥, which in turn reduces the total cost by 27.03×10 3 ¥. At the same time, the absorbing capacity of renewable energy has been increased by 86.73%, and the carbon emission has been reduced by 98.33%, which has improved the absorbing capacity of renewable energy, and achieved the goal of near-zero carbon emission while maintaining a high energy balance in the system. According to Figure 13, it can be seen that the load that can be reduced and the load that can be shifted can affect the net cost. It can be seen that with the increase of the proportional coefficient of demand response, the operating cost of the system is gradually reduced. When the proportional coefficient of the demand response of the shiftable load is 0.3 and the proportional coefficient of the demand response of the curtailable load is 0.2, the operating cost of the system is the lowest. With the gradual increase of the proportional coefficient of the translatable load demand response, the operating cost of the system also increases gradually. Therefore, an appropriate proportion of loads participating in demand response is beneficial for multitime-scale energy regulation of NZCP. However, an excessively high load demand response proportional coefficient will lead to an increase in energy regulation compensation, and at the same time, it will affect the normal operation of small industrial loads and commercial loads in the park, and reduce system benefits.   In order to fully dispatch zero-carbon resources, form a production system for carbon recycling, and achieve near-zero carbon emissions while improving system revenue. In this paper, the CCGPP-WD-P2G collaborative model is proposed on the basis of full renewable energy supply, which realizes the collaborative optimization of zero-carbon resources in the park. Four groups of scenes were proposed and compared, and the following conclusions were drawn: 1) The CCGPP-WD-P2G collaborative mode is proposed to participate in the optimal scheduling of NZCP, which makes full use of the adjustment ability of zero-carbon resources in the system and improves the economy of the system maintaining a high energy balance level. 2) In the CCGPP-WD-P2G collaborative mode, the captured CO 2 is used as the raw material for P2G gas production, which reduces carbon emissions while reducing the operating cost of P2G, and provides gas source support for CCGPP, realizing the recycling of CO 2 .
3) The flue gas storage device and the CO 2 solution storage device are respectively introduced to decouple the flue gas treatment link, the carbon capture link and the power generation link, and the renewable energy output is shifted in time and space. The load peak-valley curve is effectively stabilized, the renewable energy consumption capacity is improved, and the energy supply pressure on the upper energy grid is reduced. 4) A reasonable proportion of interruptible loads and transferable loads can be used to participate in scheduling, which can effectively reduce the net cost of system operation and improve the flexibility of energy supply and demand in the system.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusion of this article will be made available by the authors, without undue reservation.