Microsoft Word - CET--006.docx CHEMICAL ENGINEERING TRANSACTIONS VOL. 59, 2017 A publication of The Italian Association of Chemical Engineering Online at www.aidic.it/cet Guest Editors: Zhuo Yang, Junjie Ba, Jing Pan Copyright © 2017, AIDIC Servizi S.r.l. ISBN 978-88-95608- 49-5; ISSN 2283-9216 Research on the Risk Assessment of Hazardous Chemical Supply Chain and the Application Hualei Ju Shaanxi Academy of Governance, Xi’an 710068, China jhljuhualei@126.com Hazardous chemical has the characteristics of easy explosion, toxicity, corrosiveness etc. These characteristics make dangerous chemical not only causing harm to the human body, but also polluting the environment. With the development of the global economy, the application of hazardous chemical is more and more widely. Compared with the traditional supply chain, the hazardous chemical supply chain has more potential factors and faces more risks. These potential risks threaten life and property safety. This is also a great challenge to the management of hazardous chemical, especially the management of the hazardous chemical supply chain. In order to assess the risk of hazardous chemical supply chain, we propose an improved AHP algorithm in this paper. This algorithm combines the PSO algorithm with AHP to improve the traditional AHP algorithm. After that, we apply the improved AHP algorithm to assess the risk of the hazardous chemical supply chain and get good results. 1. Introduction The safety of hazardous chemical has become an international problem (Terwoert et al., 2016). Due to the continuous upgrading of the industrial chain, the trend of international industrial transfer is strengthening. In the process of production, transportation and storage, there are many risks in dangerous chemical supply chain due to the particularity of the goods. Using the thought of supply chain management to guide the hazardous chemical supply chain management brings a new opportunity for the hazardous chemical (Caroline, 2013). But we need to be aware that there are still many risks in the hazardous chemical supply chain. In order to ensure the safe operation of hazardous chemical supply chain, reduce the loss of personnel and cost, we must assess the risk of dangerous chemical supply chain. In the links of production, storage, transportation, use and waste disposal, there is no negligence of the security risks. For the management of the hazardous chemical supply chain, we need not only the supervision of relevant departments, but also the cooperation among different disciplines. Compared with other supply chain, the risk management of hazardous chemical supply chain has the particularity of risk type, risk identification and so on (Börjeson et al., 2015; Fu et al., 2016; Jiang et al., 2016; Shi et al., 2016; Tan, 2016; Zheng et al., 2016; Ji and Lan, 2016; Wang and Hao, 2016; Tao, 2016.). In the present study, many scholars have studied the risk assessment of supply chain, but the risk assessment of hazardous chemical supply chain is less. However, the risk of dangerous chemical supply chain is more extensive and it results in more serious consequences. Because of the particularity of hazardous chemical, the risk identification and the control of the hazardous chemical supply chain is a common topic of the whole society. AHP is a very practical method, which combines qualitative description and qualitative analysis. At present, AHP is widely applied in chain risk assessment (Dong et al., 2016), economy development level evaluation (Duan et al., 2016), supplier selection (Dweiri et al., 2016) and so on. PSO is proposed from the simulation of bird hunting behavior search strategy (Lin et al., 2015). In PSO, the solution of each optimization problem is called particle. Each particle has its own position, velocity and the fitness value. The particle swarm follows the current optimal particle to search in the solution space. At present, the PSO method has been applied in many fields (Dong et al., 2017; Kesavan et al., 2016; Tian et al., 2017). In this paper, we propose an improved AHP method to assess the risk of hazardous chemical supply chain. This method combines AHP with PSO method so as to assess the risk of the hazardous chemical supply DOI: 10.3303/CET1759188 Please cite this article as: Hualei Ju, 2017, Research on the risk assessment of hazardous chemical supply chain and the application, Chemical Engineering Transactions, 59, 1123-1128 DOI:10.3303/CET1759188 1123 chain better. The first part of this paper is the introduction. The first part mainly introduces the research background of this paper. The second part is the risk assessment of hazardous chemical supply chain. In the second part, we set up the risk assessment indexes of dangerous chemical supply chain. The third part introduces the PSO algorithm. The fourth part is the improved AHP algorithm. In this part, we propose an improved AHP algorithm. The fifth part is the experiment and the sixth part is the conclusion. 2. The risk assessment index of dangerous chemical supply chain Assessing the risk of hazardous chemical supply chain is a complex system assessment. The assessment involves more content and the factors will be considered more. Scientific and reasonable index of risk assessment of hazardous chemical supply chain not only related to the quality of the hazardous chemical supply chain evaluation work, but also the safety management of the operation of the hazardous chemical unit. The establishment of indicators must be reasonable and can fully reflect the impact of hazardous chemical supply chain of all factors. The risk assessment index of dangerous chemical supply chain is shown as the Table 1. Table 1: The risk assessment index of dangerous chemical supply chain First grade index Second grade index Third grade index Dangerous chemical supply chain risk Contingency management Hazard identification Monitoring and early warning Emergency plan Resource allocation Emergency rescue Emergency recovery capability Risk source index Quantity of dangerous goods Hazardous substance toxicity Inflammable and explosive dangerous goods Dangerous goods operating pressure Vulnerability index Distance from densely populated areas Population density Environmental sensitivity Logistics risk Transportation risk Storage risk Delivery risk Enterprise risk Enterprise safety level Comprehensive management evaluation Staff quality Working environment Outsourcing risk Strategic risk Supplier selection risk Other risks Technical risk Natural disaster Policy risk Reputation risk 3. PSO The space is n dimensions. The position of i(i=1, 2,…, n) is xi=(xi1, xi2,…, xiN)T and the vector is xi=(xi1, xi2,…, xiN)T. The update formulas of the speed and the position are ( 1) ( ) 1 1 2 2 ( ) ( ) k k k k id id id id k k gd id v v c r p x c r p x       (1) ( 1) ( 1)k k k id id id x x v     (2) 1124 k is the iteration number. p k id is the individual extremum of the i-th particle at k-th iteration. p k gd is the global extremum of the i-th particle at k-th iteration. c1 and c2 are acceleration constant. The rang of r1 and r2 is (0, 1). They are the independent functions. Figure 1: The flow chart of PSO 4. Improved AHP We first establish the evaluation index structure model. This model includes destination layer, criterion layer and scheme layer. After establishing the evaluation index, we construct the judgment matrix Ra={aij}nm. 11 12 1 21 22 1 2 { } n n ij n m n n nn a a a a a a Ra a a a a               (3) The weights are wi(i=1, 2,…, n). Where 1 0, 1 n i i i w w    (4) According to the judgment matrixRa, we know , 1, 2, , ; 1, 2, , ik i k a w w i n k n   (5) At the same time, the judgment matrix Ra meets 1 ij i j a w w  (6) 1 ki ik k i a a w w  (7) ( ) ( ) ki kj i k k j i j ij a a w w w w w w a    (8) 1125 According to the judgment matrixRa, we can get the weights of each evaluation index wi(i=1, 2,…, n). If the judgment matrix satisfies the formula 5, the decision maker can measure it accurately. The judgment matrix Ra has complete consistency 1 1 ( ) ( ) , 1, 2, , n n ik k i k k i i i a w w w w nw i n       (9) 1 1 [ ( )] 0 n n ij k i k k a w nw       (10) In general, the judgment matrixRa does not have complete consistency in real life. If the judgment matrix does not satisfy the consistency, we have to make some adjustments to the judgment matrix until it meets the consistency. The left of formula 10 shows the consistency of the judgment matrix. If it is close to 0, the consistency of judgment matrixRa has a high degree of the consistency. Therefore, the consistency of the judgment matrix is transformed into a linear optimization problem with constraints 1 1 min ( ) ( ) n n ik k i k k CIF n a w nw n      (11) 1 . . 0, 1, 2, , , 1 n k k k s t w k n w     (12) CIF(n) is the coherence function and the weights are wi(i=1, 2,…, n). When the judgment matrix Ra is completely consistent, the formula 11 takes the global minimum CIF(n)=0. According to the constraint condition, the global minimum is unique. Then, we use the improved PSO algorithm for consistency checking, that is, to optimize CIF(n). When CIF(n)is smallest, the judgment matrix has the highest degree of consistency. At the same time, CIF(n)is the fitness of the PSO. 1 ( 1) 1 1 1 1 1 1 1 1 2 2 2 2 1 [( ) ] [ ( ) ( )] max [( ) ] [ ( ) ( )] max k k k k k k id id f i i l id id l id id k k k k f i i l gd id l gd id iter v wv c c c r p x p x Iter iter c c c r p x p x Iter                           (13) 1 1k k k id id id x x v     (14) Where c1i and c1f are the initial value and the final value of c1 respectively. c2i and c2f are the initial value and the final value of c2 respectively. Improved AHP algorithm steps are as follows. Step 1. Construct the index structure model. Step 2. Construct the judgment matrix. Step 3. Determinate the objective function ( )CIF n . Step 4. Weight values wi(i=1, 2,…, n) are used as the optimization variable to encode the real value. The number of digits is determined by the dimension of the weights. Among them, each parameter value plus restrictions to ensure that the parameters are in the range of (0, 1). And they must meet the normalization conditions. Step 5. Population is initialized. The population represents the sets which meet the judgment matrices. The real coded strings which are generated randomly represent a set of parameter values. Step 6. The fitness is calculated according to the objective function to test the consistency of the judgment matrix. Step 7. Iterative optimization is operated by improved PSO algorithm. If the precision requirement is met or the number of iterations is reached, the iteration is terminated. Step 8. The evaluation results are obtained according to the index weight. 1126 5. Experiment In the experimental stage of this paper, we assess the hazardous chemical supply chain risk of four companies. Through the improved AHP algorithm, we first calculate the weights of each index which meet CIF(n)0.001. The weights of each index are shown in the following table. Table 2: The weights of each index Second grade index Weights Third grade index Weights Contingency management 0.21 Hazard identification 0.15 Monitoring and early warning 0.19 Emergency plan 0.13 Resource allocation 0.11 Emergency rescue 0.23 Emergency recovery capability 0.19 Risk source index 0.26 Quantity of dangerous goods 0.21 Hazardous substance toxicity 0.25 Inflammable and explosive dangerous goods 0.31 Dangerous goods operating pressure 0.23 Vulnerability index 0.19 Distance from densely populated areas 0.46 Population density 0.27 Environmental sensitivity 0.27 Logistics risk 0.14 Transportation risk 0.33 Storage risk 0.34 Delivery risk 0.33 Enterprise risk 0.12 Enterprise safety level 0.18 Comprehensive management evaluation 0.16 Staff quality 0.15 Working environment 0.13 Outsourcing risk 0.10 Strategic risk 0.16 Supplier selection risk 0.12 Other risks 0.08 Technical risk 0.38 Natural disaster 0.27 Policy risk 0.14 Reputation risk 0.21 After that, we can get the hazardous chemical results supply chain risk assessment of four companies. The results are shown as the Table 3. Table 3: The evaluation results of four companies Company A B C D evaluation results 0.7682 0.7053 0.7412 0.7829 Therefore, the ranking of the four companies is DACB 6. Conclusion The emergence of chemical has brought a great help to the development of society. When people enjoy the convenience of chemical, the safety of hazardous chemical is in sight. The emergence of the supply chain management has promoted the development of supply chain. Apply the supply chain management to the chemical supply chain can bring great convenience. However, there is a huge security hidden in the hazardous chemical supply chain due to the special nature of chemical. Therefore, it is necessary and urgent to assess the risk of hazardous chemical supply chain. In this paper, we propose an improved AHP algorithm to assess the risk of hazardous chemical supply chain. The assessment results show that the method is correct and effective. The main contents of this paper includes (1) the research background of this paper is introduced (2) the evaluation index of is proposed (3) An improved AHP method is proposed and we use this method to assess the risk of dangerous chemical supply chain. 1127 References Börjeson N., Gilek M., Karlsson M., 2015, Knowledge challenges for responsible supply chain management of chemicals in textiles – as experienced by procuring organizations, Journal of Cleaner Production, 107, 130-136, DOI: 10.1016/j.jclepro.2014.03.012 Caroline E.S., 2013, Reducing hazardous chemicals in consumer products: proactive company strategies, Journal of Cleaner Production, 44, 105-114, DOI: 10.1016/j.jclepro.2012.12.005 Dong Q.X., Cooper O., 2016, An orders-of-magnitude AHP supply chain risk assessment framework, International Journal of Production Economics, 182, 144-156, DOI: 10.1016/j.ijpe.2016.08.021 Dong R.T., Xu J.P., Lin B., 2017, ROI-based study on impact factors of distributed PV projects by LSSVM- PSO, Energy, 124, 336-349, DOI: 10.1016/j.energy.2017.02.056 Duan Y., Mu H.L., Li N., Li L.L., Xue Z.G., 2016, Research on Comprehensive Evaluation of Low Carbon Economy Development Level Based on AHP-Entropy Method: A Case Study of Dalian, Energy Procedia, 104, 468-474, DOI: 10.1016/j.egypro.2016.12.079 Dweiri F., Kumar S., Khan S.A., Jain V., 2016, Designing an integrated AHP based decision support system for supplier selection in automotive industry, Expert Systems with Applications, 62, 273-283, DOI: 10.1016/j.eswa.2016.06.030. Fu H.Y., Chen W., Song H., Peng H.W., 2016, Coordination of the supply chain of seasonal products with buy- back contract under weather-sensitive demand, Advances in Modelling and Analysis A, 53(2), 18-33. Ji H.Y., Lan C.F., 2016, On coordinating a loss-averse oem supply chain with random supply and demand, Chemical Engineering Transactions, 51, 667-672, DOI: 10.3303/CET1651112 Jiang M.L., Lin X.W., Ren H.P., 2016, Research on synergetic pricing strategy of cloud-closed loop supply chains, Advances in Modelling and Analysis A, 53(1), 30-46. Kesavan E., Gowthaman N., Tharani S., Manoharan S., Arunkumar E., 2016, Design and implementation of internal model control and particle swarm optimization based PID for heat exchanger system, International Journal of Heat and Technology, 34(3), 386-390, DOI: 10.18280/ijht.340306 Lin T., Wu P., Gao F.G., Yu Y., Wang L.H., 2015, Study on SVM temperature compensation of liquid ammonia volumetric flowmeter based on variable weight PSO, International Journal of Heat and Technology, 33(2), 151-156, DOI: 10.18280/ijht.330224. Shi H.R., Liu K.J., Peng H.H., Yao Y., 2016, Fuzzy TOPSIS-based supply chain optimization of fresh agricultural products, Advances in Modelling and Analysis A, 53(1), 186-203. Tan J., 2016, Agricultural products supply chain coordination：price and green level, Chemical Engineering Transactions, 51, 931-936, DOI: 10.3303/CET1651156 Tan J., 2016, Research on quality management of supply chain with multi manufacturer competition, Chemical Engineering Transactions, 51, 925-930, DOI: 10.3303/CET1651155 Tao J., 2016, Research on the supply chain risk prediction based on the improved intelligent algorithm, Chemical Engineering Transactions, 51, 349-354, DOI: 10.3303/CET1651059 Terwoert J., Verbist K., Heussen H., 2016, An Intervention Study on the Implementation of Control Banding in Controlling Exposure to Hazardous Chemicals in Small and Medium-sized Enterprises, Safety and Health at Work, 7, 185-193, DOI: 10.1016/j.shaw.2015.12.002. Tian S.W., Wang C.M., Zhang Z.M., 2017, A hybrid method of debris flow velocity estimation based on empirical equation, International Journal of Heat and Technology, 35(1), 147-152, DOI: 10.18280/ijht.350120 Wang Y., Hao H.X., 2016, Research on the supply chain risk assessment of the fresh agricultural products based on the improved toptsis algorithm, Chemical Engineering Transactions, 51, 445-450, DOI: 10.3303/CET1651075 Wei C., Bai Y., Chuan L., 2016, Partner selection mechanism of cooperative innovation in project-based supply chain based on the effect of knowledge level, Advances in Modelling and Analysis B, 59(1), 146- 163. Zheng X.Y., Zhu P.P., Chen S.J., 2016, Research on optimization of closed loop supply chain integrated model based on low carbon, Chemical Engineering Transactions, 51, 967-972, DOI: 10.3303/CET1651162 1128