#CO₂ Transport and Storage

#Summary

US-REGEN explicitly models the transport and storage of captured CO₂ on a regional basis. Carbon can be captured in a wide range of applications, including power plants, fuel conversion plants, certain industrial plants, and direct air capture plants. For long-term geologic storage, the captured CO₂ can be transported to injection sites and stored underground reservoirs, such as saline aquifers. The costs for transport and storage of captured carbon include investments in both pipeline capacity and CO₂ injection capacity, as well as ongoing operating and maintenance costs. Regional storage capacity is limited based on a national study with SCO₂TPRO (EPRI, 2023bopen in new window).^[1] However, investments in inter-regional CO₂ pipeline capacity can be made to access capacity in neighboring regions. Additionally, intra-regional pipeline capacity is needed to connect capture sources (e.g. power plants) with injection sites.

#CO₂ Storage Costs and Capacity

When CCS technologies are deployed in US-REGEN, captured CO₂ can be stored in regional saline formations. To access these formations, investments must be made in injection capacity and O&M costs (fixed and variable) must be paid on an ongoing basis. For this reason, storage costs are represented by independent supply curves for capital and fixed O&M costs. A variable O&M cost of $0.70/tCO₂ stored (2023 USD) is assumed across all formations. US-REGEN includes reference CO₂ supply curves and two extreme supply curves. The Reference supply curve includes reference engineering assumptions and uses the geologic characteristics reported in the SCO₂TPRO database. The Pessimistic supply curve includes engineering assumptions that lead to larger costs and geologic characteristics that are worse than reported. The Optimistic supply curve includes engineering assumptions that lead to smaller costs and geologic characteristics that are better than reported. These three supply curves capture the full range of CO₂ storage availability and costs and thus are appropriate for evaluating the sensitivity of CCS deployment to CO₂ storage assumptions. Figure 1 illustrates the estimated CO₂ storage capacity and levelized cost of CO₂ storage throughout the continental U.S. for the three storage scenarios.

As a linear programming model, US-REGEN cannot include nonlinear continuous supply curves. Rather each supply curve must be represented as a linearly increasing step function in which the first step represents the block of storage capacity with the lowest cost and each subsequent step represents blocks with increasing costs. To reduce complexity, the supply curves are represented by five steps (i.e., CO₂ storage resources) with increasing costs. Only resources available for a capital cost of less than $250 per tCO₂ per year of injection capacity are included in the supply curves.

Capital costs associated with the 5th, 25th, 50th, and 75th percentiles across the supply curves in all states and scenarios were identified and used to assign individual storage resources to the five resource classes. Using this approach, individual resource classes represent uniform capital cost ranges across all scenarios and states. Finally, the capacity-weighted average capital cost, capacity-weighted average fixed O&M cost, and total storage capacity within each resource class, state, and scenario were calculated to translate the continuous supply curves to the individual steps representing each storage resource class. Figure 2 provides an illustration of the translation between raw supply curves and modeled resource classes in the state of Colorado for the lowest cost 20 Gt of capacity.

The capital cost, fixed O&M cost, and storage capacity for each resource class in each state and scenario are provided to US-REGEN where they are aggregated into the user-defined multi-state regions. The capacity aggregation is done by summing the capacity in each resource class and scenario across the region’s member states. The capital and fixed O&M costs are aggregated by calculating the capacity-weighted averages for each resource class and scenario. Figure 3 illustrates the supply curves translated to levelized costs of CO₂ storage for US-REGEN’s default regional aggregation. A regional supply curve is provided for each storage scenario (black lines) as well as the cost and capacity of regional CO₂ storage as modeled in US-REGEN before this project (red line). The levelized cost is derived by multiplying the capital cost by a capital recovery factor of 12% and adding the fixed and variable O&M costs.

While only the capital and fixed O&M costs are provided as inputs to US-REGEN, the levelized cost of storage is a useful metric for comparing differences in storage costs across regions and scenarios. The first notable insight from Figure 3 is the significant increase in the levelized cost of storage relative to the previous version of US-REGEN (red lines). Even in the Optimistic scenario, levelized costs of storage are between 40% and 300% larger. The second notable insight is that there is little difference between the Optimistic and Reference scenarios in most regions with at least 10 Gt of storage capacity available for $6-7/tCO₂. The exceptions are MISO-North, Ohio Valley, South Atlantic, Mid-Atlantic, and New York. In particular, New York and Ohio Valley have very little capacity below $10/tCO₂ even in the Optimistic scenario. In the Pessimistic scenario, storage costs increase significantly in most regions and some regions have no economic storage capacity (e.g., New York, South Atlantic, Ohio Valley, and MISO-East). However, California, Pacific, Texas, Southeast, and MISO-South retain significant capacity below $9/tCO₂ even in the Pessimistic scenario. New England has no available CO₂ storage capacity in any scenario.

#CO₂ Transport Costs

REGEN represents CO₂ transport costs via pipeline at two levels, similar to other commodities: inter-regional CO₂ transport, for example to move captured CO₂ from a region with limited or no injection capacity to another region with better injection potential; and intra-regional CO₂ transport to connect sources of CO₂ capture with injection sites within a region. For inter-regional transport, pipeline capacity investments costs are based on a 42-inch diameter pipeline with 20% average utilization (a relatively conservative assumption, higher utilization would lead to lower costs), scaled by the distance between region centroids. For intra-regional CO₂ transport, each capture application includes a scaled pipeline investment component based on assumed flow rate and distance between capture plant and potential injection site. For carbon capture retrofits of existing power plants and new bioenergy plants with capture, siting near injection capacity will be more difficult (i.e. driven by the existing plant's location or the biomass feedstock source), thus REGEN assumes longer pipeline distances for these applications, which are assumed to be met through a hypothetical network combining feeder and trunk lines. For new gas-fired power generation and hydrogen or ammonia production, or for direct air capture, REGEN assumes siting could be more easily dictated by a suitable injection location and thus would entail shorter pipeline distances, essentially feeder only. Figure 4 summarizes the pipeline scale assumptions and associated costs for each CO₂ capture application. Figure 5 summarizes the assumed capital costs and CO₂ nominal flow rate for different pipeline sizes. Both inter-regional and intra-regional pipeline capacity also has an annual carrying cost (i.e. fixed operating and maintenance) estimated at 2.5% of capital costs.

Note that there are economies of scale with pipeline transport, i.e. the levelized cost of transport declines with increasing flow. Moreover, there is significant spatial variation both within and across regions in terms of alignment between capture plant and injection sites. The current US-REGEN estimates reflect representative average levelized costs, whereas actual transport costs for individual projects could vary considerably.