System dynamics is a simulation method that uses variables such as inventory variables, flow variables, auxiliary variables and internal feedback loops to analyze nonlinear behavior in complex systems. It was created by Forrester who is from the Massachusetts Institute of Technology in the 1860s (Forrester, 1961). Since then, system dynamics has been widely used in researches to solve the characteristic of high complexity in the system, especially via explaining the causal relationship between variables in the internal feedback loop to analyze the rules of variation in the complex system. It means the system dynamics method would analyze issues from the internal perspective rather than the outside perspective.
The relationship between variables in the system dynamics model is connected by a feedback loop. There are two types of feedback loops: positive feedback loop and negative feedback loop. In the feedback loop, any changes in one variable would lead to variations in the entire system. The positive feedback loop indicates that the increase or decrease trend of any variables in the loop will drive the system to change in the same direction, while the negative feedback loop is just the opposite.
Constructing a system dynamics model and analyzing the internal variation rule of a complex system is mainly divided into two steps. The first step is to clarify the causal relationship in the system under real conditions and establish a qualitative causality conceptual model. Afterwards, the causality diagram should be formed.
The second step is based on the established causality conceptual model, changing the existing variables into horizontal variables, auxiliary variables and other variables that can be quantified to form a stock-flow diagram. Afterward, simulation analysis would be implemented with the help of system dynamics software such as Vensim. Finally, the simulation results can be obtained.
In recent years, the system dynamic method has been extensively applied to the C&D waste management field. Yuan et al. (2011) considered the dynamics nature in the C&D waste chain to analyze the cost- benefit of C&D waste management. Yuan & Wang (2014) established a system dynamic model to determining the C&D waste disposal charging fee. Jia et al. (2017) gave a dynamic incentive mechanism for C&D waste management based on the penalty and subsidy. Mak et al. (2019) predicted the optimum waste disposal charging fee in Hong Kong by constructing a system dynamic model. Liu et al. (2020) analyzed the C&D waste recycling industry from a systematic perspective which provided a deep understanding of C&D waste recycling industry chain related issues.
Based on previous studies, the C&D waste recycling management can be divided into three parts, namely, C&D Waste Recycling, C&D Waste Directly Dumping and C&D Waste Illegal Dumping. Therefore, the conceptual model in this study for C&D waste recycling under the incentive policy consists of three feedback loops, shown in Fig. 2(a).
Loop 1: Once the amount of recycled C&D waste increases, the overall economic and environmental benefits will be improved and more favorable factors can appear to promote the development of C&D waste recycling industry. With the gradual increase of driving factors, the effectiveness of recycling regulations can also be strengthened under the general trend of waste recycling, so more effective and feasible policies can lead the development of the C&D waste recycled materials market. Hence, the market will be well developed, thereby it will effectively reduce the recycling fee. Once the recycling fee is reduced, the cost of C&D waste recycling will also be reduced. Therefore, the enterprise would like to send more C&D waste to recyclers for waste recycling. Finally, the proportion of recycled C&D waste will eventually increase. Hence, this loop is a positive feedback loop.
Loop 2: When the amount of illegal-dumping C&D waste increases, the damage to both society and the environment will increase too. The government will be no doubt to increase the supervision level and set more restraints on waste illegal-dumping in order to protect social interests. Therefore, the effectiveness of recycling regulations implementation would be improved under the attention of the government. With effective governmental supervision, the probability of being caught in illegal dumping will be greatly increased, which will cause enterprises to pay a greater price. For profit-oriented enterprises, it is obviously not acceptable. In this way, enterprises will gradually avoid choosing illegal-dumping, thereby reducing the amount of illegal-dumping waste. Therefore, this loop is a negative feedback loop.
Loop 3: Dumping C&D waste directly is the simplest and cheapest way for enterprises. When the amount of directly dumped C&D waste increases, the government will inevitably issue more restrictions to ask enterprises to take recycling measures. Therefore, the number of factors restricting the C&D waste directly dumping would increase and it will promote the effectiveness of recycling regulations implementation. As a result, the cost of directly dumping will increase. When the cost of directly dumping rises, it will encourage enterprises to carry out waste recycling measures and reduce the proportion of directly dumping waste. Ultimately the amount of directly dumped waste will reduce. Therefore, this loop is also a negative feedback loop.
Based on the above three feedback loops, the causal loop diagram is shown in Fig. 2(b). The arrow symbol in the diagram represents the causal relationship between the two connected factors, and the trend of arrow-head factor will change with the arrow-tail factor. The “+” sign indicates that the changing trend of the arrow-head factor is consistent with the variation trend of the arrow-tail factor. It means the arrow-head will increase with the arrow-tail increases or the arrow-head will decrease with the arrow-tail decrease. The “-” sign indicates that the changing trend of the arrow-head factor is opposite to the changing trend of the arrow-tail factor. The latter increases and the former decreases, and when the latter decreases, the former increases.
After the preparation, the stock-flow diagram can be drawn which is able to simulate the process according to the causal loop diagram. The stock-flow diagram is shown in Fig. 2(c).
In China, GDP and construction acreage are often used to predict the amount generated C&D waste (Chen & Yuan, 2017; Liu et al., 2014). These data come from Shanghai Statistics Yearbook from 2010 to 2018.
Table 1
GDP and construction acreage in Shanghai
Time
(Year)
|
GDP
(100 Million)
|
Construction acreage
(1000 square meter)
|
2018
|
32679.87
|
47577.35
|
2017
|
30133.86
|
41197.49
|
2016
|
26688
|
36019.72
|
2015
|
24964.99
|
36631.16
|
2014
|
23560.94
|
34994.68
|
2013
|
21602.12
|
29148.65
|
2012
|
20101.33
|
27961.55
|
2011
|
19195.69
|
24004.25
|
2010
|
16872.42
|
22996.81
|
Based on Table. 1, the relationship between GDP and construction acreage can be obtained in Eq. (1).

According to Liu et al. (2014) and expert interviews, it is assumed that each square meter of construction acreage corresponds to 10% of square meters of C&D waste. Each square meter of C&D waste weighs about 2 tons, so the amount of generated C&D waste generated can be obtained in Eq. (2).

This paper will conduct a simulation analysis for C&D waste recycling development under incentive policies from 2015 to 2035. It assumes that the incentive policy will be officially implemented in 2021. According to the published data in the Announcement of Shanghai Solid Waste Pollution Prevention Information, the amount of solid waste reported in the city is 99.65 million tons in 2015. Chen & Yuan (2017) had found that the amount of C&D waste in Shanghai is 30% of the amount of solid waste. Therefore, it is assumed that the initial amount of C&D waste in this study is 29.895 million tons. Value of constant variables are shown in Appendix Table. A1, and other variables and table functions are also listed in Appendix Table. A2 and Table. A3 at end of the manuscript.
Model verification
In order to ensure the rationality of the given system dynamics model in this paper. This system dynamic model needs to be verified after the establishment of the system stock-flow diagram and the confirmation of the functional relationship between the auxiliary variables in the model.
The relationship between the fine and the amount of illegal dumping C&D waste is selected for verification (Jia et al., 2017; Jia & Yan, 2018). The fines are selected as 0 yuan/ton, 50 yuan/ton, 150 yuan/ton, 300 yuan/ton, 500 yuan/ton and 700 yuan/ton for simulation, respectively.
It can be seen in Fig.3, the amount of illegal dumping C&D waste gradually decreases as the amount of fines increases. This result is consistent with the reality and is consistent with the simulation results of Jia et al. (2017) and Jia & Yan (2018), which proves that the model built in this paper is realistic and feasible.