#Electricity Delivery and Storage

#Inter-Regional Transmission

The electric model allows for inter-regional transfers of power using a bubble-and-spoke representation. In each segment or dispatch period, the amount of power transfer between a pair of regions is limited by the total transmission capacity between them.^[1] Existing inter-regional transmission capacity is derived from transfer capacity values between transmission areas as reported by ABB Velocity Suite. We have also developed regional transfer capacity values derived from National Renewable Energy Laboratory's ReEDS model (NREL 2017) and EPA's IPM 5.13 model (EPA 2017) for testing the sensitivity of model results to transmission assumptions. Figure 1 shows the existing power transfer capacity between the default 16 regions. Power flows between regions are subject to line losses proportional to the centroid-based distance between the two regions. For every 100 miles of long-distance transmission, US-REGEN assumes 0.5% power losses. US-REGEN allows additional transmission capacity to be installed between regions at a cost of $2.5M per GW-mile (in year 2023 dollars), based on an average of recent projects.

The electricity transmission system is complex in terms of planning, operation, and organization. In many scenarios, including the reference scenario, additional build limits are imposed to reflect difficulties with siting or public acceptance. These limits have been generated in conjunction with experts from EPRI's transmission planning area and also reflect recent experience with the rate of transmission capacity additions. The constraint takes the form of an upper bound of 15% increase in transmission capacity on each link per five-year time step. Under this constraint, links with more existing capacity can add more future capacity, and links with limited existing capacity (such as the AC-DC interties between interconnects) are restricted to smaller absolute additions. Over the 30 year period between 2020 and 2050, the maximal increase in transmission capacity on any given link under this growth constraint is a factor of 2.3. In the current version of the model, international imports and exports of electricity are held constant over time at base year levels.

#Intra-Regional Transmission and Distribution

While inter-regional transmission capacity and dispatch is modeled explicitly, most of the delivery system for electricity entails intra-regional transmission and distribution. US-REGEN does not explicitly resolve the electricity network below the level of model regions, but it does include estimated costs for expanding and maintaining the grid. These estimates are based partly on the extent of wind and solar deployment, but more generally they scale with electricity demand from the end-use sectors.

New wind and solar capacity incurs a one-time charge of $315 per kW and $125 per kW respectively (in year 2023 dollars) to reflect the incremental transmission investment required to incorporate generally remotely located wind and solar capacity within a region. This number is based on a combination of discussions with internal and external transmission experts and a review of the relevant literature (for example see Mills et al. (2009), Jaske (2010)). There is no added transmission charge for repowered wind (i.e. new wind turbines replacing existing turbines at existing sites). An incremental distribution charge for distributed PV is included as described in the Distributed Energy Resources section.

General expenditures on transmission and distribution within a region are captured as a component of the retail price for delivered electricity. The levelized costs of delivery from the wholesale power market to end-use sectors are based on observed differences in wholesale and retail prices in current EIA data. These costs are summarized in Table 1, denominated in both kWh and MMBtu for comparison with non-electric fuel delivery costs.

#Storage

US-REGEN includes several technology options for electricity storage. The model endogenously chooses investment levels and dispatch of storage on an hourly basis. Storage investment, charge, and discharge decisions are co-optimized with other technology investment and dispatch decisions. By default, US-REGEN represents four possible storage technology classes, and other technologies can easily be added. These storage classes include pumped hydro (existing capacity only), lithium ion batteries, compressed air, and a generic bulk storage technology that could reflect a flow battery, liquid air storage, or other longer-duration option. Each technology is characterized by a charge penalty, which represents the ratio of input to output energy, or roundtrip efficiency. Note that compressed air has has a charge penalty less than 1 because it also takes a natural gas input (4.38 MMBtu, or 1.28 MWh, gas per MWh electricity output) that converts at a loss to electricity. Each technology also has an assumed charge loss rate and an assumed lifetime for new and existing builds. In some configurations, the model also includes thermal storage embedded in a concentrated solar plant, where the costs for the power block, solar field, receiver, and storage components are modeled separately. However, in this case CSP storage can only charge from its own field.

US-REGEN's assumptions for storage cost and performance parameters are based on EPRI's latest storage technology assessment (EPRI, 2024aopen in new window), which includes a comparison of multiple analyses and specifies cost ranges for several storage technologies. Electricity storage cost projections include costs of the system, grid integration equipment, fixed and variable maintenance, and system decommissioning. In US-REGEN, electricity storage costs are represented as a linear function of both nominal output or power capacity (the "door"), expressed in $/kW, and nominal energy capacity (the "room"), expressed in $/kWh, so the total cost projections from the report are disaggregated into these "door" and "room" costs for use in the model. Unless exogenously fixed for a given technology, the model can endogenously choose the ratio of energy capacity to output capacity, i.e. the duration of the system. In general, lithium ion batteries have the highest "room" costs, lowest "door" costs, and lowest roundtrip losses (the equivalent of variable costs for storage), making them most suitable for shorter duration (i.e. frequently cycled) applications. On the other hand, bulk storage and compressed air storage (as well as hydrogen-based power-gas-power storage) have lower "room" costs, higher "door" costs, and are less efficient, which implies they would likely be only cost-effectively deployed for longer duration (i.e. infrequently cycled) applications. Table 2 shows cost and performance assumptions for each storage technology. Capital costs are shown for 2030 and 2050 (in year 2023 dollars). The model also considers lower and higher sensitivities for cost declines over time.

US-REGEN's dynamic formulation uses an aggregation of representative hours into load segments to significantly speed up computation. This creates challenges for representing storage effectively, as chronology is lost by construction, but can be approximated using a state-space approach. This approach allows storage to be included in the intertemporal optimization as part of a long-term investment planning scenario. Nonetheless, the static rental model with full 8760 hourly resolution offers the best assessment of the economic potential of electricity storage and trade-offs among alternative storage and generation options.