Network analysis of pore structure of coral reef limestone and its implications for seepage flow

Determining the seepage characteristics of coral reef limestone is highly significant for engineering activities on relevant strata. In this paper, the pore structure information of two types of coral reef limestones with distinct pore characteristics was obtained by combining computed tomography (CT) scanning technology and pore network modeling. We regarded the pore structure of coral reef limestone as a complex network, and the seepage inside is analogous to information transmission. Network analysis was performed on this network considering both global and individual network indicators. It was shown that the pore network of coral reef limestones has a short seepage distance and strong connectivity, and exhibits a prominent clustering effect. The centrality of network nodes indicated that very few pore nodes have extremely high centrality. Based on the Dijkstra algorithm, the shortest percolation paths along three orthogonal directions were obtained and visualized for both samples. This showed that the shortest percolation paths always converge to a few paths in all directions for a pore network of coral reef limestone with coarse pores. There existed a preferential seepage path characterized by more intensive shortest percolated paths. Some pore nodes with higher betweenness centrality were located at a crucial location in a few shortest percolation paths. A new indicator VB considering both the volume and betweenness centrality of the pore node was proposed. Those pore nodes with large VB values were identified as key pore nodes in the network, as seepage simulations showed that blockage of these key nodes causes a dramatic decrease in permeability.