Research on Carbon Reduction Paths in Recycling Industrial Parks based on System Dynamics

: The "double carbon" target is a medium-to-long-term national strategy proposed by China to combat climate change. The industrial sector is one of the key areas for the implementation of the "double carbon" target. Therefore, studying the association between carbon emission factors and carbon emissions is crucial to reduce greenhouse gas emissions from industrial activities. In the present study, the association between factors affecting carbon emissions and carbon emissions in a circular economy were investigated for an industrial park in Northwest China. A carbon emission system dynamics model for this circular economy industrial park was constructed, in reference to the relevant national policies and the current condition of the park. Five different scenarios were utilized to dynamically simulate the impact of rapid economic development, energy restructuring, industrial restructuring, and technological development, and carbon emission reduction paths for industrial parks were explored. The results showed that, the park would peak at 1134.67 thousand tons of CO 2 in 2032, according to the baseline scenario, with industrial energy consumption accounting for over 80% of the total emissions. A combined regulation scenario, with increased investment in research and development and environmental management, would achieve a peak in 2030, with a relatively lower peak of 1062.88 thousand tons of CO 2 . Our findings provides new insights into the paths of carbon emission reduction in recycling industrial parks.


Introduction
With the socioeconomic development of the human population, fossil fuels have become an important cornerstone of industrial and economic activities [1]. Greenhouse gas emissions, represented by carbon dioxide (CO 2 ), from industrial activities have led to global climate change, which in turn has created certain pressures and challenges to sustainable development. As the largest global carbon emitter, China has a great responsibility of addressing climate change [2][3]. Liu et al. (2018) explored the timing of carbon peaking in China using a coupled STIRPAT model and a system dynamic model, and used the simulation results to help the government propose carbon reduction policies [4]. Chen et al. (2022) constructed a system dynamics model and combined it with scenario analysis to dynamically simulate the carbon emissions of marine fisheries in the northern marine economic zone of China's marine fisheries development region [5]. Zhang et al. (2022) constructed a system dynamics model to analyze the peak carbon emissions of urban water systems under different regional water demands in China [6]. Guo et al. (2022) used the system dynamics method to predict the peak of vehicle carbon emissions in Chinese cities and put forth relevant suggestions [7].
A circular industry requires reasonable arrangement of the symbiotic coupling of industrial, chains, realization of multilevel utilization of materials and energy recycling in industrial production, and scientific development based on the principles of clean production and requirements of circular economy development. Based on the characteristics of a circular industrial park, the present study selected the relevant factors affecting carbon emission and simulated them with a system dynamics model to provide a scientific basis for the park to assume a low-carbon emission path and achieve the goal of the "double carbon" target.

Subjects and boundaries of research
The research object of the present study was a mineral resource-rich recycling industrial park in Northwest China, where a circular economy system has been established with power generation; coal chemical industry, nonferrous metals; iron, and steel as the mainstays and building materials, fertilizers, new materials, new energy and biotechnology as the downstream-products.
All greenhouse gas emissions from production bases and facilities are under the operational control of enterprises in the park. Emission sources can be divided into direct and indirect offers, and further subdivided into indirect emissions from energy sources and other indirect emissions. The scope includes emissions from fossil fuel combustion, emissions from alternative fuel combustion, emissions from non-biomass carbon of exhaust species, emissions from the decomposition of raw materials, and CO 2 emissions from net purchased electricity and heat use.

Cause and effect flow chart of carbon emissions
Based on the study of the park and related studies, the major factors affecting carbon emissions of the recycling industrial park were determined, including economic development, energy structure, industrial structure, primary energy use, number of employees, technological innovation, and environmental management level [8]. Based on these factor, a carbon emission prediction model of the studied recycling industrial park was designed with four subsystems: economic development, energy use, population and environment, and science and technology innovation.
Based on the abovementioned influencing factors, the causal loop of the recycling industrial park carbon emissions system was drawn using system dynamics software (see Fig.1).

Carbon emissions stock flow map
In the present study, we used the carbon emission data of a circular industrial park in Northwest China from during 2013-2019 and other related data to establish a model and create a flow chart of the carbon emission system model of the circular economy industrial park with Vensim 9.3.2. Next, we quantitatively analyzed each variable and used the system dynamics simulation software to evaluate and correct the model for obtaining simulation results closer to the actual carbon emissions. The flow chart of the carbon emission system of this circular industrial park is shown in Fig.2 5. Rate of change in industrial energy consumption= LN (Per capita production value of the park) /82*LN (Recycling rate) *(-1)-0.12^ (LN (Recycling rate) *(-1)) (Unit: DMNL). 9. Change in carbon emissions per unit of energy= Industrial structure impact factor* Science and Technology R&D Impact Factor* Energy structure impact factor /5* (-1) (Unit: thousand tons of CO 2 / thousand tons of standard coal).

Scenario simulation scheme setting
In the present study, a scenario simulation method combined with a system dynamics model was used for simulation [9]. The impacts of different scenarios on the system results were obtained by changing the various parameters of the factors and setting different scenarios for simulation. Based on the original model prediction scenario as the base scenario, the economic growth rate regulation scenario, R&D investment regulation scenario, environmental management investment regulation scenario, energy and industrial restructuring scenario, and comprehensive regulation scenario were set. Five variables, namely the growth rate of total output value of the park, proportion of R&D investment, proportion of investment in environmental management, effect of industrial structure, and effect of energy structure, were used as the primarily control variables in the scenario analysis, the specific scenarios are summarized in Table 1. The growth rate of total output value of the park is reduced by 2%, the proportion of R&D investment is increased by 10%, the proportion of investment in environmental management is increased by 7%, the impact factor of industrial structure is reduced by 3%, and the impact factor of energy structure is reduced by 5%.

Data sources
In this study, to maintain data consistency, all data were obtained from on-site research, enterprise-related information (2013-2019), and local statistical yearbooks of park.

Model validity check
In the present study, based on the research data of the target park and validity tests used in related studies, the historical test method was selected to validate the model, using historical data of 2013-2019. The model was considered valid if the relative error was within 15% [10]. Based on the above validity test, the relative error between the simulated and actual values of the five selected variables was within 10%. Therefore, the carbon emission system dynamics model developed in the present study is reliable and could effectively simulate carbon emission changes in the park (see Table 2).

Model dynamic simulation results
Based on the results of simulation, carbon emissions from the park will continue to increase between 2020 and 2030, reaching a peak of 113,467 kiloton CO 2 in 2032 and decreasing to 34,809 kiloton CO 2 in 2060 (see Fig.3a). The percentage of industrial energy consumption in carbon emissions from different emission sources will decrease from 87% to 48%, while the percentage of industrial process carbon emissions will continue to increase. Carbon emissions from purchased electricity will peak at 8,904 kiloton CO 2 in 2033, followed by a decrease to 2,105 kiloton CO 2 in 2060 (see Fig.3b). From the result of systematic simulation, the change trends of total carbon emissions and industrial energy consumption carbon emissions in the five scenarios were consistent with the change trends of the baseline scenario. In the simulation results of the economic growth rate regulation scenario, the impact of the growth rate of the total output value of the park on the peak of carbon emissions was small. The peak time of carbon emissions was consistent with the baseline scenario, both in 2032, with a peak value of 113,361 kiloton CO 2 . After peaking, the carbon emissions of the park would continue to decline, and by 2060, they would drop to 33,532 kiloton CO 2 , being 3.7% lower than those in the baseline scenario.
In the simulation results of the R&D input regulation scenario, the R&D input ratio produced a greater impact on the total carbon emissions and industrial energy consumption carbon emissions. The carbon emission peak of the park advanced to 2031, reaching 107,587 kiloton CO 2 , indicating a 5.2% decrease compared with emission in the baseline scenario. Meanwhile, industrial energy carbon emissions would peak at 9,2421 kiloton CO 2 , indicating a 10.9% decrease on average compared with emissions in the baseline scenario.
In the simulation results of the environmental management investment regulation scenario, the percentage of environmental management investment only affected the carbon sink and CO 2 utilization of the park green area. The carbon emissions of the park will peak at 113,163 kiloton CO 2 in 2032, indicating a 0.3% decrease compared with emissions in the baseline scenario, and will further decrease to 34,025 kiloton CO 2 by 2060.
In the simulation results of the integrated regulation scenario, the carbon peak of the park advanced to 2030, reaching 106,288 kiloton CO 2 , indicating a 6.3% decrease compared with emissions in the baseline scenario. By 2060, the carbon emissions of the park will drop as low as 22,734 kiloton CO 2 , the CO 2 utilization of the park will be 7,607 kiloton CO 2 , the total output value of the park will be 0.5% lower than that in the baseline scenario, and the carbon emissions will peak in the same period. Under this scenario, the total output of the park will decrease by 0.5% compared with that in the baseline scenario, and the peak carbon emissions will decrease by 6.3% during the same period (see Fig.4).

Conclusions and discussion
In the present study, Vensim 9.3.2 was used to investigate the impacts of various factors on carbon emissions of the recycling industrial park, and the reliability of the model was verified based on historical data of the target site. On this basis, five different scenarios were established to dynamically simulate the trends of carbon emissions of this park from 2020 to 2060. Under the base scenario, industrial energy consumption carbon emissions accounted for 87% of the total carbon emissions, and the proportion of coal exceeded that of other fossil energy sources accounting for 72%. Under the R&D input share regulation scenario, industrial energy consumption carbon emissions decreased by 7% compared with those in the base scenario. Under the comprehensive regulation scenario, industrial and energy restructuring can reduce carbon emissions per unit of energy by 15.9%. To limit effects on the normal production and development of the park, reducing the proportion of coal use and optimizing the energy structure by increasing the proportion of clean energy will allow for the diversification of the energy structure of the park , rendering it more sustainable.
Under the environmental management investment regulation scenario, the carbon sink and carbon dioxide utilization in the park would be respectively 8% and 12% higher than those in the baseline scenario. In the development process of a recycling industrial park, enterprises with carbon dioxide processing and utilization capacity should be introduced according to their own advantages; waste from different enterprises should be fully utilized, and the association between waste and energy conversion should be established in the park to promote the synergistic development of pollution and carbon reduction while increasing economic benefits. The green space system is the major component of the ecological environment of the park, which is key to regulate the balance of carbon and oxygen and forms a basis for the sustainable development of the ecological environment.
In summary, future carbon emissions from industrial parks show an overall increasing trend. However, if the management of industrial parks increases the proportion of clean industries and the enterprises in industrial parks continuously optimize their energy structure, increase investment in science and technology, and formulate more comprehensive emission reduction strategies, carbon emissions of industrial parks will continuously decline, and achieving the emission reduction goals of industrial parks would be easier.