Microsoft Word - 02Revised.doc


 CHEMICAL ENGINEERING TRANSACTIONS  
 

VOL. 55, 2016 

A publication of 

 
The Italian Association 

of Chemical Engineering 
Online at www.aidic.it/cet 

Guest Editors: Tichun Wang, Hongyang Zhang, Lei Tian
Copyright © 2016, AIDIC Servizi S.r.l., 
ISBN 978-88-95608-46-4; ISSN 2283-9216 

Logistics Network Construction of Hazardous Goods Based 
on Analytic Hierarchy Process 

Xiuli Tang 

School of Logistics, Beijing Wuzi University, Beijing, China 

 
Chemical hazardous goods logistics has become a new branch of logistics, but how to choose the location of 
logistics center and build the logistics network is still an urgent problem to be solved by logistics enterprise. 
Based on this, this paper studies the logistics enterprise chemical hazardous goods logistics network 
construction. First of all, construct the logistics center location model of hazardous chemicals, establish 
decision framework of logistics center location, and give priority weight of each decision criterion; then, this 
paper uses two stage analysis method to study chemical hazardous goods logistics network logistics 
construction. In the first stage, the relationship between logistics center, inventory, transportation and service 
quality are defined; in the second stage, the main task is to adopt analytic hierarchy process (AHP) to 
determine the evaluation criteria of logistics network used in this study, and design a logistics network 
construction model of chemical hazardous goods. 

1. Introduction  
Hazardous chemicals are mostly flammable, explosive, toxic, and corrosive, which leads to the security, 
environmental protection and other issues in the production, sales, transportation, warehousing and other 
aspects. Thus, a little negligence will bring about risks and threats to production enterprises, consumers, and 
social security, and cause great damage to the environment. At present, the chemical hazardous goods 
logistics in China is still in the preliminary stage of development, and the logistics layout structure also has 
defects, which result in that chemical transportation accidents frequently occurred in China. In consequence, 
to establish suitable chemical logistics network is particularly important (Deng et al., 2013). Based on this, this 
paper, on the basis of analytical hierarchy process, studies the procedures, measures and steps to construct 
the logistics network of hazardous chemicals, strives to make it clear the relationship between the various 
factors of the construction of chemical hazardous goods logistics network, and develops and builds a set of 
logistics network construction process adapt to the development of the logistics enterprises, so as to 
determine that the logistics network can be good in supply, fast in turnover, and low in consumption, cost, thus 
ensure production effectively and smoothly carried out. 

2. Research contents and methods 
2.1 Construction of logistics center location model for chemical hazardous goods 
In this paper, the AHP method is adopted as a decision-making model to discuss the importance of the 
selection of the decision-making criteria for the logistics center of the chemical hazardous goods. The decision 
model of logistics center location based on AHP is shown in Figure 1. 

                               
 
 

 

 
   

                                                  
DOI: 10.3303/CET1655078

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Please cite this article as: Tang X.L., 2016, Logistics network construction of hazardous goods based on analytic hierarchy process, Chemical 
Engineering Transactions, 55, 463-468  DOI:10.3303/CET1655078   

463



 

Figure 1: Decision model 

After establishing the AHP decision model of the logistics center of chemical hazardous goods, the relative 
priority weight of each criterion is calculated by the expert to fill in the AHP questionnaire. The pairwise 
comparison problem in the questionnaire is composed of the texture comparison and the criteria comparison 
in the SuperDecisions software. Immediately after the experts fill in the questionnaire, the questionnaire data 
is input into the SuperDecisions software, understand whether each matrix meets the requirement of 
consistency (C.R.). If do not meet the requirement of consistency (C.R.≦0.1), then after illustrated and 
discussed with experts, fill a 2-2 value, and at the same time amend its consistent values, to meet the 
requirements of AHP. Summarize the pair comparison matrix obtained when taking each decision texture and 
decision criteria as the major considerations, and sort out the priority weight and overall priority weight under 
various decision criteria textures, as shown in Table 1, in which, items with "*" represent the most important 
criteria under the texture.  

Table 1: Priority weight for decision criteria 

TextureDecision criteria 
The priority weight under the 
texture 

The overall priority 
weight 

I 

1. Operator’s innovation* 0.43331 0.181113 
2. Sales channel capability 0.08539 0.035714 
3. Capacity of production process planning 0.08087 0.033896 
4. Strategic planning and overall positioning 0.24511 0.103301 

II 

5. Personnel training plan and education training 0.15292 0.063898 
6. Market intelligence feedback 0.23613 0.061328 
7. Application of information technology 0.12721 0.033050 
8. Construction of job information 0.11782 0.031435 

III 

9. Establishment of customer response system* 0.51604 0.134972 
10. Circulation processing equipment 0.10325 0.001588 
11. Dangerous goods storage and transportation 
equipment 

0.02517 0.000422 

12. Production scale economies 0.44632 0.006823 

IV 

13. Stability of supply source 0.42376 0.006493 
14. Land expansion 0.30214 0.093021 
15. Local government policy support * 0.59477 0.182484 
16. Main traffic facilities approach 0.10219 0.031342 

 
From analysis of Table 1, it is known that the first 3 most essential chemical hazardous goods logistics center 
selection criteria are the local government strategy support (0.1825), operator’s innovation (0.1811), and 
establishment of customer response system (0.1350), whose importance are in more than 9%. At the same 
time, in the first nine important decision-making criteria, there are four criteria belonging to the texture of 
"Management team concept and ability", the weight of a total of about 38%, which suggests that the texture is 
of great importance (Giannetti et al., 2013). The top three key decision criteria are discussed and specifically 
illustrated combined with the actual situation of establishing a logistics center of hazardous chemicals: 1) The 
local government strategy support: logistics center building, hazardous goods special storage and 

464



transportation equipment, added value and so on facilities construction cost are very expensive, so private 
companies cannot independently complete the overall construction. Therefore, it is one of the essential factory 
that enterprise need to consider about that whether can strive for the support of government, acquisition tax 
incentives, land rent concessions and other government supports in the location of logistics center of chemical 
hazardous goods. 2) Operators’ innovative: mainly refers to that the enterprises need to innovate in 
technology, products and services, and business model. 

2.2 Study on the construction of logistics network of chemical hazardous goods 
The purpose of logistics network construction is: to build the logistics network taking pursuing the minimum 
logistics cost and the maximum customer satisfaction as the starting point, and from the angle of integrating 
the existing logistics resources. In this section, we first discuss the relationship between logistics center, 
inventory, transportation and service level, and then design a multi-objective planning model. 

2.2.1 Relationship between logistics center, inventory, transportation and service level 
The construction of the logistics network is not only to consider the district or number of the logistics center 
itself, it also needs to consider the relationship between information strategy, inventory strategy and transport 
strategy, as shown in Figure 2. The transportation cos and inventory management are the most frequently 
quantitative discussion topics. 

 

 

Figure 2: Construction of the relationship between the elements of logistics network 

From Figure 2, we can see the relationship between the four kinds of networks. But this study suggests that 
although the information processing occupies an important position in the logistics network construction, in the 
actual planning, the information processing mechanism can be included in the setting cost of logistics center 
to quantify. While the customer service level has a clear response in the logistics network construction cost, 
and when the decision makers construct the logistics network, the consideration focus is usually on the 
service level the total cost of logistics and logistics network of customer (Guerin 2015). Therefore, this 
research mainly aims at the logistics network construction, and then carries on the detailed explanation to the 
relationship between the logistics center and the inventory network, and that between transportation network 
and service level. 
1) The relationship between logistics center and inventory management 
In the case of giving the shortage rate and the total demand, the quantity of safety stock reflects a linear 
relationship with the quantity of warehouse (DC), so the effect of increasing the warehouse on the average 
inventory can be shown in Figure 3. 
 

 

Figure 3. The relationship between average inventory and logistics center warehouse quantity 

465



2) The relationship between the logistics center and the transportation system 
Transportation costs often occupies a large proportion in the overall logistics costs, so how to design a most 
efficient transport system is the problem that enterprises must pay attention to in the logistics network 
construction. The relationship between transportation cost and warehouse number is shown in Figure 4. 

 

Figure 4: The relationship between transportation cost and the quantity of logistics center warehouse  

From the analysis of Figure 4, it is known that, when the enterprise warehouse number is quite small, the 
customer orders must be processed with direct shipping, so the transportation cost is very high; but if the 
enterprise increase warehouse number, economic law can be used to combine with shipment and provide 
logistics services with transportation method. At the same time, the total cost of transportation can be reduced 
with the increase of warehouse; and when the warehouse number is too much, the number of consolidated 
shipments will decline to not able to use the scale of transportation economy law, which will lead to excessive 
increase in the total cost of transportation. 
3) The relationship between logistics center and service level 
Generally speaking, the relationship between the number of logistics center warehouse and customer service 
level is shown in Figure 5. 
 

 

Figure 5: The relationship between the service level and the quantity of the logistics center warehouse  

Figure 5 shows that, with the increase of the quantity of the warehouse, the relative cost of logistics also 
increases, and service level is improved; when the warehouse number is increased to a certain amount, if it 
continues to increase, the logistics cost will substantially increase for too large scale of logistics center. And 
once the number of warehouse is too much, it will result in that the transmission of information is not smooth, 
leading to reduction of customer service level.  

2.3.2 Multi-objective programming model construction of logistics network  
1) Based on the role of the new Austrian group in the distribution of methanol product, this paper discusses 
the construction of the logistics network, including the main elements: i methanol processing plant, k customer 
demand point and j potential logistics center location. 
2) The mode of transportation of the product includes the product transporting from the processing factory to 
the logistics center, and then from the logistics center to the demand points, or directly transporting from the 
factory to the customer these two kinds. 
The goal formula of the model total cost is shown in the Eq (1). Minimize: 

466



j
ji j k

ikjk
i k

ikik

j k
jkjk

i j
ijij

i
ii

j
jj

Shzyuzc

ybxaFFTCZ

δλ

δδ




Ω+++++

+++==

)()(

1

   (1) 

The target function that the customer is satisfied with is shown in Eq (2). Maximize:  

  +==
k

k
i i j

jkik
j

jkjk dwzwyCSRateZ )(2    (2) 

The parameters of decision variables are set as follows (Hao, et al., 2015): 
xij suggests the transport volume of processing factory i to logistics center j; yjk indicates the transport volume 
from logistics center j to customer k; zik refers to the transport volume directly transported from the processing 
plant to the customer k; δj represents binary variables for whether logistics center j is open, and δj=1 indicates 
open, while δj=0 suggests not open; δi represents binary variables for whether the processing factory i opens 
direct transportation, δi=0 indicates open, and δi=1 suggests not open; wjk is the service level rate of customer 
k served by the logistics center j; wik suggests the service level rate of customer k served by the factory i. 
The parameters of this research model are set as follows: 
Si suggests the supply of the processing plant i; dk is customer k demand; wj indicates the minimum shipments 
of logistics center j; Wj is the maximum capacity of logistics center j; Wi is the minimum shipments of the 
processing factory i; Wi represents the maximum capacity of the processing plant i; Fj is the setup cost of 
logistics center j; Fi suggests the setup cost of the processing plant setting up direct distribution; u is the unit 
handling cost for goods transported to customers (Netro, et al., 2016); u said the unit treatment cost; aij is the 
unit transportation cost of the processing plant i to logistics center j; bjk indicates the unit transportation cost 
from logistics center j to the customer k; cik suggests the unit transportation cost form the processing plant to 
the customer k. S is the logistics center safety inventory, Ω is the average inventory holding rate, λ is the 
demand for a market, and the number of logistics center inventory is S+Ω·λ. 
This study only takes the network and environment two qualitative criteria as the second stage evaluation 
criteria. The main reason for planning in the first stage of planning, is to consider the quantified criteria in the 
model. As a result, do not consider the cost criterion. This study adopted logistics network evaluation criteria is 
shown in Table 2. 

Table 2: Logistics network qualitative factor evaluation criterion 

 Evaluation criteria Explanation 

Network 

Transportation speed (time) The time it takes to transport to the point demanded. 

Transportation stability 
It refers to the time difference for transporting same quantity 
goods to the same demanded point. 

Traffic convenience 
It suggests the road traffic condition of the place where the 
facilities are located in. 

Market proximity 
It indicates whether the place where the facilities are located in 
is close to the demand side or the supply side. 

Rationality of the scope of service
It is the rationality of districts where the facilities are 
responsible for service. 

Inter facility mutual support It is the degree that facilities support each other. 

Environment
Time difference of information It represents the speed of information transmission. 
Applicability of laws and 
regulations 

It is the degree of overall planning in line with laws and 
regulations. 

3. Research results and discussion  
This research aims at the problem of logistics network construction, and puts forward the planning method of 
two stages. In this two stage method, it includes qualitative and quantitative criteria so that the overall planning 
is more in line with the needs of enterprises. In the first phase, for the criteria quantified, use multi-objective 
programming method to solve the advantages and disadvantages of several groups; in the second stage, 
based on the combination of the above solution, add the needed to consider qualitative criteria for multi-
criteria evaluation. Sort the various programs so as to select the best plan. The new Austrian group chemical 
hazardous goods logistics center project is in the planning stage. According to the development of enterprises 

467



and the progress of the project, design chemical hazardous goods logistics network construction flow chart of 
the new Austrian group, as shown in Figure 6.  

 

Figure 6: Evaluation flow chart 

4. Conclusion  
The site selection of logistics center and the construction of logistics network is a new problem in the study of 
logistics in China. But the accident of hazardous goods distribution is a typical "low probability while serious 
consequence" event. Once it occurs, the loss of life and property and the environmental impact are very 
serious. Therefore, the AHP decision model is taken as decision model, to select decision criteria affecting 
logistics center location of hazardous chemical, and then construct the AHP decision framework of logistics 
center location, so as to find out the decisive factors which should be noticed in the determination of logistics 
center location (Qian et al., 2015). And then, taking to determine the location of logistics center as the premise, 
analyze the relationship between logistics center, inventory transportation and service level, and design multi-
objective programming model for logistics equipment construction. Combined with the actual requirements of 
the construction of the logistics network of the chemical hazardous goods, sort out the qualitative factors 
evaluation criteria, design model of chemical hazardous goods logistics network logistics enterprises, to 
provide support for the possibility of future logistics network construction of hazardous chemical, which has 
certain practical significance. 

Acknowledgments  

Project funds support: Beijing Nova Program (Z121106002512043); The General Social Science Program of 
Beijing Municipal Education Commission (SM201310037001); Philosophical and Social Science Program 
(12JGC107); Beijing excellent talent training Program (2013D005009000001) 

Reference  

Deng A.M., Zhong H.F., Li H.L., Wang J., 2013, Review of Researches on Logistics Industry Clusters 
Sustainable Development. InInternational Asia Conference on Industrial Engineering and Management 
Innovation (IEMI2012) Proceedings, Springer Berlin Heidelberg, 1227-1235. 

Giannetti B.F., Bonilla S.H., Almeida C.M., 2013, An emergy-based evaluation of a reverse logistics network 
for steel recycling. Journal of cleaner Production, 46, 48-57. 

Guerin T.F., 2015, Understanding the Causes of Spills from the Supply and Handling of Chemicals at 
Resource Construction Sites: A Case Study.Remediation Journal, 25, 3, 115-145. 

Hao Q., Zhang F., Liu Z., Qin L., 2015, Design of chemical industrial park integrated information management 
platform based on cloud computing and IOT (The Internet of Things) technologies. International Journal of 
Smart Home, 9, 4, 35-46. 

Netro Z.G.C., Álvarez J.E.M., Carrillo A.C., Flores R.G., 2016, Solid waste management in Mexico’s offshore 
platform construction: determining potential supply for a reverse logistics process. NETNOMICS: 
Economic Research and Electronic Networking, 1-24. 

Qian Y., Juncheng J., Hanhua Y., Bureau J.P.S.S., 2015, Study on Graphic Demonstration Technology for 
Emergency Decision-making in Event of Highway Accidents Involving Hazardous Chemical Materials. 
Logistics Technology, 17, 012. 

468