⚙️ How-To Wind turbines Maps & visualisation Notebook By

How to Map Wind Power Capacity by County

On this page

Problem statement

American wind power is concentrated: a few hundred rural counties in the plains, the Texas panhandle, the passes of California and the Columbia Gorge carry most of it. To map that, each turbine's rated capacity is summed by its county and joined to county boundaries. The US Wind Turbine Database already records the county of every turbine as a five-digit FIPS code, so the work is a groupby and a join, with three things to watch: the code must be read as text, a few thousand turbines have no capacity, and some codes belong to counties that no longer exist. This page maps version 9.1 by county.

Quick answer

Group the turbines by t_fips (read as text), sum t_cap (kilowatts) into megawatts, and merge onto the 2024 county boundaries by GEOID:

c = wt.groupby("t_fips").agg(turbines=("case_id", "size"), mw=("t_cap", lambda s: s.sum() / 1e3))
m = counties.merge(c, left_on="GEOID", right_index=True, how="left")
print(m.nlargest(6, "mw")[["NAME", "STUSPS", "turbines", "mw"]].round(0).to_string(index=False))
print(f"counties with turbines: {len(c)}; joined to a 2024 county: {m.mw.notna().sum()}")
    NAME STUSPS  turbines     mw
    Kern     CA    3250.0 3254.0
   Nolan     TX    1349.0 2173.0
 Willacy     TX     680.0 1688.0
 Gilliam     OR     736.0 1612.0
Torrance     NM     963.0 1595.0
Converse     WY     645.0 1585.0
counties with turbines: 721; joined to a 2024 county: 717

Turbines stand in 721 counties, and 717 of the codes match a 2024 county. Kern County, California, has the most capacity, 3,254 MW from 3,250 turbines: Tehachapi Pass, with its old small turbines and new large ones side by side. The next five are in Texas, Oregon, New Mexico and Wyoming, each with 1,585 to 2,173 MW from far fewer turbines.

County map of the lower 48 shaded by wind capacity, darkest in Kern County, west Texas, Oklahoma, Kansas, Iowa, Wyoming and the Columbia Gorge; most eastern and southern counties have none. County map of the lower 48 shaded by wind capacity, darkest in Kern County, west Texas, Oklahoma, Kansas, Iowa, Wyoming and the Columbia Gorge; most eastern and southern counties have none.
Rated wind capacity by county.

Summing, joining and the counties that changed

1. Turbines without a capacity

unknown = wt.t_cap.isna()
print(f"turbines without a capacity: {unknown.sum():,} ({unknown.mean():.1%}); summed capacity {wt.t_cap.sum() / 1e6:.1f} GW")
print(wt[unknown].groupby("t_state").size().nlargest(4).to_string())
turbines without a capacity: 5,056 (6.5%); summed capacity 159.6 GW
t_state
CA    1604
NM    1004
TX     739
CO     422

5,056 turbines, 6.5% of the file, have no capacity, so the county totals are a floor. They are not spread evenly: California has 1,604 of them and New Mexico 1,004, so their counties are understated most. A county map of turbine counts, which needs no capacity, is the check.

2. Codes that do not join

gone = c[~c.index.isin(counties.GEOID)]
print(gone.join(wt.groupby("t_fips").t_county.first()).to_string())
        turbines     mw           t_county
t_fips                                    
02270         14  2.000           Kusilvak
09005          2  5.700  Litchfield County
09009          1  0.100   New Haven County
66010          1  0.275               Guam

Connecticut's eight counties were replaced by nine planning regions in the Census Bureau's 2022 boundaries, so codes 09005 and 09009 exist only in older files. Alaska's Wade Hampton Census Area became Kusilvak in 2015 and got a new code, 02158; the turbine file has the new name under the old code. 66010 is Guam, which the county file leaves out. Only 18 turbines and 8.1 MW are affected, so they can be mapped by hand or left out.

3. Concentration

s = c.mw.sort_values(ascending=False)
share = s.cumsum() / s.sum()
print(f"counties holding half of the capacity: {(share < 0.5).sum() + 1}; 90%: {(share < 0.9).sum() + 1}; of {len(s)}")
counties holding half of the capacity: 87; 90%: 306; of 721

Half of the country's wind capacity is in 87 counties, and 90% in 306 of the 721 counties that have any. Most US counties have no wind turbines at all.

Cumulative curve of the share of US wind capacity against counties ranked by capacity, reaching 50% at 87 counties and 90% at 306 of 721. Cumulative curve of the share of US wind capacity against counties ranked by capacity, reaching 50% at 87 counties and 90% at 306 of 721.
Half of the capacity is in 87 counties.

Capacity per area, and a map

The data used on this page

The US Geological Survey's US Wind Turbine Database, version 9.1 of 28 September 2026 (Hoen and others, doi:10.5066/F7TX3DN0), public domain, saved on this site on 19 October 2026 because USGS replaces the file every quarter, and the Census Bureau's 2024 cartographic boundary counties (1:20 million).

from pathlib import Path
import zipfile

import geopandas as gpd
import pandas as pd
import requests

Path("data").mkdir(exist_ok=True)
# USGS US Wind Turbine Database v9.1 (28 September 2026) with its metadata and changelog, Census 2024 cartographic
# boundary counties and the EIA-860 2025 wind table. Saved 19 October 2026: USGS replaces the file every quarter.
# Live: https://energy.usgs.gov/uswtdb/assets/data/uswtdbCSV.zip
ARCHIVE = "usgs-uswtdb-v9-1-2026-10-19.zip"
if not Path(f"data/{ARCHIVE}").exists():
    Path(f"data/{ARCHIVE}").write_bytes(requests.get(f"https://data.spatialworkflow.io/{ARCHIVE}", timeout=600).content)
Z = zipfile.ZipFile(f"data/{ARCHIVE}")
wt = pd.read_csv(Z.open("uswtdb_V9_1_20260928.csv"), dtype={"t_fips": str})
Path("data/cb_2024_us_county_20m.zip").write_bytes(Z.read("cb_2024_us_county_20m.zip"))
counties = gpd.read_file("data/cb_2024_us_county_20m.zip")
print(f"US Wind Turbine Database v9.1: {len(wt):,} turbines; counties: {len(counties):,}")
US Wind Turbine Database v9.1: 77,839 turbines; counties: 3,222

Example 1: megawatts per 1,000 km²

m["km2"] = m.to_crs(5070).area / 1e6
m["mw_per_1000km2"] = m.mw / m.km2 * 1000
print(m.nlargest(5, "mw_per_1000km2")[["NAME", "STUSPS", "mw", "km2", "mw_per_1000km2"]].round(1).to_string(index=False))
        NAME STUSPS     mw    km2  mw_per_1000km2
      Benton     IN 1249.2 1053.1          1186.3
Santa Isabel     PR  101.2   92.7          1092.3
       Nolan     TX 2173.0 2366.4           918.3
       White     IN 1202.6 1321.7           910.0
     Willacy     TX 1687.7 1886.2           894.7

A large county can hold a lot of capacity and still be thinly covered. Divided by land area, Benton County, Indiana, ranks first, and Santa Isabel in Puerto Rico second, on the strength of a single 101 MW project in a small municipality.

Horizontal bars of megawatts per 1,000 square kilometres for the twelve most densely covered counties, led by Benton, Indiana, at 1,186 and Santa Isabel, Puerto Rico, at 1,092. Horizontal bars of megawatts per 1,000 square kilometres for the twelve most densely covered counties, led by Benton, Indiana, at 1,186 and Santa Isabel, Puerto Rico, at 1,092.
Capacity per area: the most densely covered counties.

Example 2: the choropleth

away = ["02", "15", "60", "66", "69", "72", "78"]
lower48 = m[~m.STATEFP.isin(away)].to_crs(5070)
ax = lower48.plot(column="mw", scheme="FisherJenks", k=6, cmap="YlGnBu", legend=True, figsize=(11, 6),
                  missing_kwds={"color": "#eeeeee"}, edgecolor="white", linewidth=0.1)
ax.set_axis_off()
print(f"lower 48 counties with turbines: {lower48.mw.notna().sum()}")
lower 48 counties with turbines: 695

The map shows 695 counties of the lower 48 with turbines. Fisher-Jenks breaks suit a distribution this skewed, where most counties have a little and a few have more than a gigawatt. The dark band runs from west Texas up through Oklahoma, Kansas and Iowa, with separate dark patches in Wyoming, the Columbia Gorge and Kern County.