Empirical Study Of Refined Oil Supply Network Based On Complex Network Theory

Refined oil is an important energy to national economic development. The market of it strongly influences the operation and development of the entire national economy. Especially after China's reform and opening-up, and after WTO accession, the requirement for opening domestic refined oil market's has brought greater pressure on the oil market, conditions of which have drawn wide attention.The object studied in the paper is in the most economically developed province in western China. The refined oil supply chain there divides into 4 levels: CNPC, CPC and social oil enterprises as the first level, the municipal branches of these 3 enterprises as the second level, specific regional distributors as the forth level and finally the gas stations, which involves 3096 market entities, relatively a huge number.Paper draws on the complex network theory to research the refined oil supply network in one province on a macro-level. And build a refined oil supply weighted network with 3268 nodes and 8217 edges. In this network, the node is each main market entity, two nodes have a common edge if corresponding market entities have trading relation and weights the edges by the sales. On the basis of that paper goes on analyzing the external environmental factors of affecting the distribution of gas station nodes and the improvement ways on the direct suppliers based on some complex network indicators. It is also the first time to introduce the supply weighted network as the object of study in the empirical field of complex networks, which has expanded the scope of empirical research to a certain extent.Paper firstly studied the non-weighted network and found that its degree distribution satisfied the power-law. So it's a scale-free network with an exponent, similar to majority of the studied real network, but suppliers and retailers nodes in this network has different statistical behavior on degree distribution. And this network has a small network diameter and average path length, but its centrality coefficient is 0, indicating that the network has a certain small-world characteristics, but it is not a rigorous "Small World" in complex network theory, that is likely to be the general network characteristics of all the supply-chain networks. Finally, the most important nodes are concentrated in three cities, the provincial capital and the other two second major cities near to the provincial capital.After that, paper researched on the weighted situation. The network is found that its node strength distribution also satisfied the power-law with an exponent, still similar to majority of the studied real network, and the power-law behavior in the node strength distribution is much strict than that in the degree distribution. Researches on the unit weight find that the wholesalers to gas stations associated relationships relative to the gas stations to the wholesalers trading relationship is broad and light, and the difference of the unit weight of the wholesaler nodes in the same level of node weight rank is huge, that is, the difference of the transactions relationship closeness of different wholesalers to gas stations is huge. On the indicator of the disparity in the weight, the gas stations which have more purchase channels then generally have small disparity in the weight, further more, the wholesalers which directly make transaction with the gas. stations are more necessary to pay attention to this indicator which could be used to balance the benefits and the transaction costs arising from its trading relationship with the gas stations. The last, this network is a disassortative network.Finally, the analysis revealed that the factors affect the distribution of the gas station nodes are mainly the secondary and tertiary industries, and the secondary industry is the most obvious. Then according to the result of the indicators of the unit weight and disparity in the weight distribution drive a moderate improvement for the third-level nodes which is the direct wholesalers to the gas stations, and the results showed that adjusting the distributions construction of the nodes satisfying the requirements within its 5% of annual sales has been able to see a more obvious target to the improve performance.The introduction of the weight increased the complexity of the network and the difficulty of the study, too. Applying the complex network theory to do the research on the supply weighted network is just a preliminary exploration. The needs to improve and further the studies are still many and urgent, in particular the need to introduce the dynamic method of the complex network theory in the research, which could provide important assistance to understand the formation mechanism, the robustness or other issues of the supply weighted network to and other issues.