Skip to content
InfraMap
Open the map
Data catalog · Model output

Modeled curtailment

Experimental InfraMap estimate of the energy US wind and solar plants likely withheld, from physical potential and the real-time price at each node.

Updated November 2, 2026

About the data

Modeled curtailment is not operator-reported curtailment. It is an experimental hourly estimate of energy a wind or solar plant likely withheld, from its physical potential and the real-time price at its node, using a response fitted on ERCOT plants.

The cohort is 1,943 plants in the seven ISO and RTO markets, 896 wind and 1,047 solar. Each needs a reviewed resource-level pricing node.

Production tax credit receipt cannot be observed, so the model shows two bounds: a lower bound with the credit and an upper bound without it. Across the covered fleet, 2025 totals 23.0 TWh with the credit and 48.4 TWh without it.

This is a model output, not observed data. InfraMap labels it as a model everywhere it appears.

Coverage
1,943 plants, 896 wind and 1,047 solar, in the seven ISO and RTO markets
Served
From January 2014, where potential and a complete real-time node price overlap
Grain
Hourly estimates; monthly totals in the plant’s local month
Fit
ERCOT dispatch reductions, January to April 2026; 250 fitted ERCOT plants use their own response

Access

Part of InfraMap Pro. See pricing.

Where it appears: Output tab, agent, MCP.

MCP tools that return it: get_project_modeled_curtailment, compare_project_curtailment, investigate_project_curtailment. The MCP server is part of Pro; setup is on MCP server.

Sources

Periods and update notes for each source are on Data sources.

Known gaps

Limits
  • An estimate, not measured curtailment. Transfer outside the fitted ERCOT plants and to other markets is unvalidated.
  • On July 1 to 28, 2026, fitted wind plants were about 27% low against the reference and fitted solar plants about 27% high.
  • Neither bound establishes credit receipt, credit election, repowering or contract terms. The range is a scenario spread, not a confidence interval.
  • Hourly mean prices omit sub-hourly dispatch, offer curves, contracts, outages and network limits.

Missing values stay missing: a null means the source has no usable record, never zero. Zero and negative values are shown as reported.

Method and validation

How the model is built, what it was tested against and where it is weak is on Methodology.

Questions it answers

  • Which ERCOT solar plants had the most modeled curtailment in spring 2025?
  • How did modeled curtailment for SPP wind change from spring 2024 to spring 2025, split into volume and rate?
  • When did a given West Texas wind farm likely curtail in March 2025, and at what prices?

Ask the analyst

Terms: Curtailment, Negative prices.