Research · published 2026-07-07 · download the county-level CSV

1 in 15 Michigan household water wells sits within a quarter mile of a documented fuel leak

Michigan does not require anyone to test a private well after it is drilled — sampled once at construction, then never again unless the owner pays for it. The state also carries one of the country’s largest backlogs of underground fuel-tank leaks it has documented but not finished cleaning up. We mapped where those two facts meet.

Comparing 1,072,031 Michigan household water wells (Wellogic) against the U.S. EPA’s UST Finder database of underground-storage-tank releases — documented leaks of gasoline, diesel, and heating oil at 23,470 Michigan sites:

• 69,323 household water wells (6.5%) lie within 1,320 ft (a quarter mile) of at least one documented petroleum release.
• 35,662 (3.3%) lie within a quarter mile of a release the state still lists as open — reported but not yet cleaned up (EPA snapshot, 2026-06-12).

None of these households is required to test their water. Public water systems near the same leaks are monitored on a federal schedule; private wells fall entirely outside that law.

EPA’s own UST Finder analysis estimated 27,535 Michigan private domestic wells within 1,500 ft of an active leaking-tank site — the third-highest count of any state — and roughly 250,000 nationally. EPA’s FY2025 performance report puts Michigan’s cleanup backlog at 7,812 open releases, second in the nation.

Counties with the most household wells near an open leak

Download the full county table (CSV)

What this does and does not say

• Proximity is not contamination. Distance from a documented release is a screening signal, not evidence any specific well is affected. Whether a plume reaches a well depends on hydrogeology, depth, direction, and remediation history. Testing is the only way to know — which is the point: no one is required to.
• What we counted. Michigan-recorded household water wells, excluding wells flagged abandoned and all monitoring, irrigation, and public-supply wells. This is the best public proxy for homes on well water, though some records are older wells since replaced or disconnected.
• “Open” status lags. The EPA release snapshot dates to 2026-06-12; states close releases every year, so the current open count is somewhat lower. The “any documented release” figure — cumulative history — is not affected.

Methodology

• Wells: 1,072,031 Michigan household water wells with coordinates from Wellogic (refreshed weekly; see our general methodology).
• Releases: 23,470 Michigan release sites with coordinates from EPA UST Finder (8,674 listed open in the snapshot).
• Join: great-circle (haversine) distance, well-to-release, threshold 1,320 ft. Each well is counted once no matter how many releases are nearby.
• Why a quarter mile: peer-reviewed work behind UST Finder used 1,500 ft as the outer bound of observed petroleum plume lengths (Weaver et al. 2017); we use a tighter 1,320 ft, so our threshold is the more conservative of the two.
• Checks: results were spot-verified at the address level (flagged wells and their nearest release share streets and ZIP codes) and screened for coordinate artifacts (no bulk co-located records).

Sources

• Michigan well records: Wellogic
• EPA UST Finder (release locations and status)
• EPA, Semiannual Report of UST Performance Measures, FY2025 (state cleanup backlogs)
• Weaver, Murray & Kremer (2017), Science of the Total Environment 609:1589–1599 (plume-length basis for the distance threshold)
• Murray, Hall, Weaver & Kremer (2021), JAWRA 57(5):828–843 (EPA’s domestic-well proximity estimates)
• ASTSWMO/EPA UST Finder briefing (state-by-state wells-near-releases counts)