关键词: 分形理论, 原油管道网络, 输油规模系数, 加权最短距离

Abstract: By the end of 2008, about 90% of China’s crude oil transportation has been taken by pipeline, and it has been the primary way of China’s crude oil flow. So that it is important to deeply analyze the spatial structure of China’s national and regional crude oil pipeline networks. On the basis of a huge mass of reference materials, this paper examines three fractal dimensions, that is, weighted length-radius, dendrite-radius and spatial correlation, and then elaborates the influencing factors of this structure, including the distribution of networks’nodes and lines, the development of regional networks and others.In sum, the amended results reveal that:1) As to national networks, the weighted length-radius dimension is characterized as linear distribution in the east while scattered distribution in the west, which reflects that the networks tend to cluster and the networks’density has become lower from the east to the west. Similarly, the dendrite-radius dimension changes from scatters to linear distribution from the east to the west, reflecting the networks’structure gets complicated and the connectivity is improved accordingly. Differed from two dimensions above, the dimension of spatial correlation expresses the whole scattered distribution,which reflects the general low accessibility. The reasons are as follows. First, the trunk of pipeline networks in East China have relatively developed, contrasting with the networks in South China and West China. Second, there are evenly some isolated pipelines in East China’s networks, which limit eastern networks’connectivity. Third, some nodes of national networks are not located at the trunk pipelines, resulting in general low accessibility. 2) As to regional networks, the overall distribution of density characteristics is similar to that of national networks, while the overall connectivity characteristic is different. Specifically, as to Xinjiang Region, its fractal dimension value of dendrite-radius is over 2, that is, its connectivity in the central part of the networks is lower than in the surrounding area, which is abnormal;the dimension value of dendrite-radius in Shanxi-Gansu-Ningxia networks is 1.216, reflecting its connectivity is relatively good, compared with less than one in East China and null value in South China. Furthermore, as to the accessibility, all regions’dimension values of spatial correlation are less than one, but the spatial correlation in East China tends to be in linear distribution. The reasons are as follows. First, the indirect oil transportation pattern of oilfield A-oilfield B-…-refineries in Xinjiang networks, increases the branches of networks, making it too complicated and limiting the connectivity of the networks. Second, the construction of pipeline networks in East China is lagging better than in other regions. Third, the construction of networks in South China is lagging behind other regions, thus its networks hasn’t been shaping. 3) With the theory and methods of fractal, it could be effective to reflect the spatial structure and its distribution of China’s crude oil pipeline networks. In the future, it is necessary to further analyze the harmony or correction between fractal feature of regional pipeline networks and regional oil production and consumption, then to reveal the interactive mechanism between them.

Key words: fractal theory, crude oil pipeline networks, oil scaling index, weighted shortest distance