(343c) Optimization of the CCS Network Under Harsh Geographical Conditions, Using GIS and MINLP Modeling

With the evermore increasing amount of greenhouse gas emissions every year, various efforts to mitigate its amount have earned worldwide attention. Carbon capture and storage, a practical technology to achieve this task, has been studied from different viewpoints, namely capture, transportation, and utilization of CO2. While the CO2 capture technology has been studied very deeply, studies for practical implementation of the CO2 transportation infrastructure has received relatively less attention. Although the CO2 capture technology is the most critical from an economic point of view, research on the topic of economic transportation of CO2 is essential for commercialization of the CCS technology, since mass amounts of it have to be transferred across long distances.
For mass amounts of CO2 transport, pipelines are deemed as the most economical option. Previous studies concerning this problem have introduced the method of pipeline network optimization regarding the geographical conditions to minimize the cost of nation-wide transport. These studies used a combination of geographical information systems and mixed-integer program (MIP) modeling to optimize the network. However, geographical conditions vary greatly according to the region of pipeline installation, and linear pipeline models are inadequate in rigorously evaluating the cost of pipelines.
In this study, a pipeline network for transporting mass amounts of CO2 in harsh geographical conditions was optimized, using mixed-integer nonlinear programming (MINLP). A commercial GIS software ArcGIS was used to obtain the least cost paths from the CO2 sources to the reservoirs, then the cost factor for penalizing the pipelines built on a certain plain was segmented into more rigorous levels. This data was transferred into GAMS, an optimization software and designed into a MINLP problem, using a weight-based cost model for evaluating the pipeline costs, for more accurate calculations. A case study in South Korea was conducted to test the effect of rigorous geographical penalization and cost model usage. The results of the study showed that a more rigorous network can be obtained by implementing more segmented penalty factors corresponding to the varying elevation. Also, it showed that rather than avoiding the rough landscape entirely, it may be more economical to gather the emitted CO2 from nearby sources into local junction nodes, then transport it through mainstream pipelines to the reservoir.