Skip to content
BelowTrace

Data & method

How BelowTrace works

BelowTrace combines public records and published program rules for one Detroit address. It never guesses where a private sewer line runs, and I label every fact by how much is actually known.

Three kinds of facts

Recorded, straight from a public record
A City, DWSD, FEMA or HUD dataset, or a published program document. I link the source and date.
Estimated, worked out from records
For example: your census block group's income share compared with the test the City uses to choose alleys. Only the program can confirm it.
Unknown, not in any public record
Where your private line runs, its condition, DWSD's camera inspection results, and the alley repair schedule. I say who can answer.

Checked live for each address

Address search

Turns your address into a location. Nothing is stored.

Esri World Geocoder (US Census Geocoder as backup)

Property details

Parcel, year built, property class, homestead, tax status.

City of Detroit, Parcels (current)

Flood zone

PSRP excludes homes in FEMA's Special Flood Hazard Area.

FEMA National Flood Hazard Layer

Area income

The Alley Sewer Repair Program prioritizes areas where more than 50% of residents are low or moderate income.

HUD Low/Mod Income by Block Group (ACS 2016 to 2020)

City open data

The repository keeps a small, reviewable source receipt for this snapshot: each public query, record count and raw-file SHA-256. The large raw downloads can be recreated from those public services. Each dataset carries the day it was pulled, the sewer and 311 snapshots are from 2026-09-18, and the permit, district and calibration data were added on 2026-09-21.

City sewer mainspulled 2026-09-18

5,220 main segments with install year, material, size and depth. Partial coverage: only mains with DWSD cleaning work orders.

DWSD sewer cleaning dashboard (gravity mains, partial coverage)

Sewer constructionpulled 2026-09-18

2,074 project segments with phase (construction, bidding, finished) and years.

DWSD Capital Improvement Projects (public view)

PSRP neighborhoodspulled 2026-09-18

The 97 neighborhoods where the Private Sewer Repair Program accepts applications.

City of Detroit, PSRP neighborhoods

Private sewer permitspulled 2026-09-21

2,722 plumbing permits for sewer, lateral, cleanout and backwater-valve work since 2019, by address. The closest public record of a private line's history, it never shows where the line runs.

BSEED trades permits (plumbing)

Council districtspulled 2026-09-21

Which district an address sits in, and how many alleys in that district are in the first round of contracts.

City of Detroit, Council districts 2026

311 reportspulled 2026-09-18

Water-in-basement investigations and cave-ins since January 2023, counted within 200 m.

Improve Detroit 311 requests

Program rules & official guidance (checked 2026-09-21)

The PSRP screener applies the City's published rules. The program guide lists an income limit of 80% of area median income on pages 3 to 4 and 50% on page 10; the 2026 policy repeats both. I show incomes between the two as “might qualify” rather than deciding for the City. Income limits are MSHDA's Wayne County limits effective May 1, 2026, for a household of four, $52,400 (50%) and $83,850 (80%).

Derived from the above (2026-09-21)

Three measurements this site makes itself. Each is reproducible by running scripts/calibrate-asrp.mjs and scripts/fetch-permits.mjs against the snapshots above. Full method and caveats.

Alley program calibration
138 alleys under contract (65 in construction, 73 in procurement) measured against 47 earlier alley projects at 500 m. Median basement-flooding reports: 18 vs 45.
District distribution
Alleys in the first round by council district: D1 55 · D2 0 · D3 0 · D4 0 · D5 0 · D6 80 · D7 3.
Private repair permits
2,722 permits, 2019 to 2026, of which 1,288 are backwater valves. Median basement-flooding reports within 500 m of a permitted address: 38.
Public record coverage
A random sample of 1,500 basement-flooding reports: 29% have a city sewer main on public record within 120 m; 71% have none.

Limits

Privacy

No accounts, no tracking. Addresses are sent to the address service and public data services to build the report, and cached on the server for up to a day so repeat lookups are fast. Screener answers never leave your browser.