To meet next generation energy needs such as wind- and solar-generated electricity, enhanced oil recovery (EOR), CO2 capture and storage (CCS), and biofuels, the US will have to construct tens to hundreds of thousands of kilometers of new transmission lines and pipelines. Energy network models are central to optimizing these energy resources, including how best to produce, transport, and deliver energy-related products such as oil, natural gas, electricity, and CO2. Consequently, understanding how to model new transmission lines and pipelines is central to this process. However, current energy models use simplifying assumptions for deploying pipelines and transmission lines, leading to the design of more costly and inefficient energy networks. In this paper, we introduce a two-stage optimization approach for modeling CCS infrastructure. We show how CO2 pipelines with discrete capacities can be linearized' without loss of information and accuracy, therefore allowing necessarily complex energy models to be solved. We demonstrate the new approach by designing a CCS network that collects large volumes of anthropogenic CO2 (up to 45 million tonnes of CO2 per year) from ethylene production facilities and delivers the CO2 to depleted oil fields to stimulate recovery through EOR. Utilization of anthropogenic CO2 has great potential to jumpstart commercial-scale CCS while simultaneously reducing the carbon footprint of domestic oil production. Model outputs illustrate the engineering challenge and spatial extent of CCS infrastructure, as well as the costs (or profits) of deploying CCS technology. We show that the new linearized approach is able to offer insights that other network approaches cannot reveal and how the approach can change how we develop future energy systems including transporting massive volumes of shale gas and biofuels as well as electricity transmission for wind and solar energy. Published 2012. This article is a U.S. Government work and is in the public domain in the USA.