Costs#
Costs and Candidate Resources#
In PyPSA-USA, candidate resource forecasted capital and operating costs are defined by the NREL Annual Technology Baseline (ATB) accessed through the PUDL project. The model currently uses the 2024 ATB which provides data for expected costs across the years 2024 - 2050 (planning horizons earlier than 2024 use the 2024 costs). The full ATB scenario grid is exported by build_cost_data for every planning horizon, and the configuration below selects which slice of that grid each carrier ultimately uses.
Selecting an ATB scenario and model case#
ATB exposes two orthogonal axes that bound the cost trajectory of every technology:
scenario—Moderate(default),Advanced, orConservative. Reflects how aggressively technology costs decline over time.Advancedis most optimistic;Conservativeis most pessimistic.model_case—Market(default) orR&D.Marketbakes current policy incentives (PTC/ITC) into the financing assumptions;R&Dstrips those out and isolates pure technology learning.
The global default applies to every carrier unless an override is set:
costs:
atb:
model_case: "Market" # Market, R&D
scenario: "Moderate" # Moderate, Advanced, Conservative
Per-carrier overrides#
To stress-test specific carriers under a different scenario without touching the global default, add an overrides block keyed by the carrier’s pypsa-name. Either field can be set independently — anything left out falls back to the global default:
costs:
atb:
model_case: "Market"
scenario: "Moderate"
overrides:
solar:
scenario: "Advanced" # cheaper solar
onwind:
scenario: "Advanced"
model_case: "R&D" # no-PTC view of onshore wind
nuclear:
scenario: "Conservative" # higher-cost nuclear
Resolution order, highest priority first:
Per-carrier override under
costs.atb.overrides[<pypsa-name>]Global
costs.atb.scenario/costs.atb.model_caseHardcoded fallback (
Moderate/Market)
Carrier names must match the keys in ATB_TECH_MAPPER (workflow/scripts/constants.py) — e.g. solar, onwind, offwind, offwind_floating, nuclear, SMR, CCGT, CCGT-95CCS, coal, coal-95CCS, geothermal, biomass, 4hr_battery_storage, etc.
How selection happens#
build_cost_data writes one row per (pypsa-name, parameter, atb_scenario, atb_model_case) into resources/costs/costs_{year}.csv, so every ATB combination is available without rerunning the rule. Scenario-independent rows (EGS supply curves, transmission costs, emissions factors) carry NA in the scenario columns and apply universally. The load_costs helper in _helpers.py reads the config and selects the right row per carrier before pivoting, so changing costs.atb only requires re-running the downstream rules — not build_cost_data.
To reflect regional differences, capital costs are adjusted using EIA state-level CapEx multipliers.
Candidate Resources#
Coal Plants: With and without Carbon Capture Storage (CCS) at 95% and 99% capture rates.
Natural Gas: Combustion Turbines and Combined Cycle plants, with and without 95% CCS.
Hydrogen Combustion Turbines: Hydrogen Combusion Turbines are implemented under the assumption of market-available hydrogen drop-in fuel. Following the default assumptions in the ReEDS Hydrogen implementation. This implementation does not account for the energy or costs required to produce or transport the fuel. Future work will implement a more detailed production, transport, and storage model of hydrogen.
Nuclear Reactors: Large Nuclear Reactors (AP1000) and Small Modular Reactors
Renewable Energy: Utility-scale onshore wind, fixed-bottom and floating offshore wind, utility-scale solar.
Battery Energy Storage: 2-10 hour Battery Energy Storage Systems (BESS).
Pumped Hydro Storage (PHS): Supply curves for 8-12 hour PHS are integrated from the NREL Closed-Loop PHS dataset.
Enhanced Geothermal Systems (EGS): Methods for implementation will be released in a forthcoming paper.
Fuel Costs#
PyPSA-USA integrates fuel costs that varry across spatial scopes and temporal scales. For more information, see here
Sector Costs#
Running sector studies will use the same power system costs as electrical only studies. Costs specific to each sector can be found in the service sector, transportation sector, and industrial sector pages accordingly.