Shaft Report
Tracking America's $11 billion bet on cleaning up 500,000 abandoned mines — one site at a time.
Channel: Shaft Report Tagline: Tracking America’s $11 billion bet on cleaning up 500,000 abandoned mines — one site at a time. Niche: Abandoned Mine Lands remediation tracking — environmental cleanup progress, contamination data, community impact, and the massive federal investment in mine reclamation. Target audience: Environmental journalists, community activists living near abandoned mines, mining history enthusiasts, policy researchers, geology/adventure buffs, and the 78 million Americans who live within 3 miles of a Superfund site. Also: anyone who’s ever fallen down a YouTube rabbit hole watching mine exploration videos and wondered “but who’s cleaning this up?” Why now: The Bipartisan Infrastructure Law allocated $11.3 billion to abandoned mine land reclamation — the largest such investment in US history. States are actively deploying funds NOW (PA: $15.2M in March 2026, NM: $20M for uranium mines in March 2026). USGS published its first comprehensive national abandoned mine inventory summary in March 2025. The money is flowing, sites are being cleaned, and nobody is tracking it in a readable way. Government dashboards are impenetrable. News covers individual stories but nobody connects the dots.
Content Example
”The River That Came Back: How $4.2 Million Turned Bennett Branch from Orange Wasteland to Trout Stream”
Published by Shaft Report — Appalachian Recovery Series
For seventy years, Bennett Branch of Sinnemahoning Creek ran the color of rust. The 55 miles of waterway cutting through Cameron and Elk Counties in north-central Pennsylvania wasn’t really a river anymore — it was a drainage pipe for a century of coal mining, carrying dissolved iron, aluminum, and sulfuric acid from dozens of abandoned mine openings into what should have been one of the finest cold-water fisheries in the eastern United States.
The numbers told the story before the science did. In 1990, water monitoring station PA-BC-08 recorded a pH of 3.2 — roughly the acidity of orange juice. Iron concentrations hit 47 mg/L, nearly fifty times the 1.0 mg/L threshold that aquatic life can tolerate. Macroinvertebrate surveys found virtually nothing. The river was, in biological terms, dead.
Then the money started arriving.
Between 1994 and 2018, a coalition of the Pennsylvania DEP, the Office of Surface Mining, watershed associations, and a small army of volunteers installed sixteen passive treatment systems across the Bennett Branch watershed — limestone channels, anoxic drains, settling ponds, and constructed wetlands that filter acid mine drainage the way a kidney filters blood. Total cost: approximately $4.2 million.
The results, measured at the same monitoring station in 2024:
- pH: 6.8 (neutral — up from 3.2)
- Iron: 0.4 mg/L (down 99.1% from 47 mg/L)
- Aluminum: 0.2 mg/L (down 97% from 6.8 mg/L)
- Macroinvertebrate taxa: 23 species (up from near-zero)
- Brook trout: confirmed reproducing naturally since 2019
The before-and-after satellite imagery is striking. Where the streambed was once coated in orange iron precipitate visible from space, the water now runs clear enough to see the gravel substrate that brook trout need to spawn.
But Bennett Branch isn’t just an environmental success story. It’s an economic one. Cameron County — population 4,500 and shrinking — has seen a quiet tourism revival built on fishing access. The Bucktail Scenic Byway now promotes the watershed. Two outfitters have opened since 2020. The county’s outdoor recreation tax revenue increased 31% between 2018 and 2024.
The Shaft Report Scorecard for Bennett Branch:
| Metric | Before (1990) | After (2024) | Change |
|---|---|---|---|
| pH | 3.2 | 6.8 | +112% |
| Iron (mg/L) | 47.0 | 0.4 | -99.1% |
| Aluminum (mg/L) | 6.8 | 0.2 | -97.0% |
| Fish species | 0 | 12 | ♾️ |
| Macro taxa | ~0 | 23 | ♾️ |
| Investment | — | $4.2M | — |
| Jobs created | — | ~35 | — |
Data sources: EPA Section 319 Success Story PA_Bennett_Branch; PA DEP water monitoring records; USGS Water Quality Portal station data; Cameron County economic reports.
What’s next: Three additional passive treatment systems are funded under the 2021 Infrastructure Law, targeting remaining acid sources in the upper tributaries. The goal: full Cold Water Fishery designation by 2028.
Every mine has a story. Every cleanup has a number. Shaft Report tracks both.
Data Sources
- USGS MRDS API — 300,000+ mine records, GeoJSON via REST (
mrdata.usgs.gov/mrds/search-bbox.php). Mine names, locations, commodities, status, geology. The backbone of the site’s interactive map. - OSMRE e-AMLIS — 24,000+ abandoned mine problem areas with hazard type, status (unfunded/funded/completed), state, cost estimates. Downloadable via DOE EDX portal (
edx.netl.doe.gov). - EPA Envirofacts API — REST API for CERCLIS/Superfund data. Links mine sites to cleanup status, responsible parties, milestones.
- EPA Water Quality Portal — REST API aggregating water monitoring data from 900+ agencies. Query by location to get water chemistry near mine sites — pH, metals, conductivity.
- EPA FRS API — Facility Registry System connecting mine sites across databases.
- USGS USMIN — Mine features from topographic maps, western US focus.
- OSMRE IIJA updates — Bipartisan Infrastructure Law AML program funding announcements, state allocations. RSS/scrape.
- State AML program press releases — Funding announcements, project completions. Scrape from state DEP/DNR sites.
- Sentinel-2 satellite imagery — Free, 10m resolution, via Copernicus Open Access Hub API. For before/after comparisons of remediated sites.
- Census Bureau API — Demographics for communities near AML sites for environmental justice scoring.
Automation Pipeline
- Schedule: Daily at 06:00 UTC (GitHub Actions cron). Full site rebuild weekly on Sundays. Breaking state funding announcements trigger additional builds via webhook.
- Collect:
- Query USGS MRDS API for mine records (incremental, by region rotation — different state each day)
- Pull latest e-AMLIS data monthly (bulk download from EDX)
- Query EPA Water Quality Portal for monitoring stations within 5km of tracked mine sites
- Scrape OSMRE IIJA page and top-10 state AML program pages for new funding/project announcements
- Fetch Sentinel-2 imagery tiles for recently-remediated sites (quarterly)
- Process:
- AI cross-references new mine records with cleanup status databases
- Identifies sites with significant water quality changes (trending better or worse)
- Matches funding announcements to specific mine sites in the database
- Generates narrative for 1-2 “site stories” per week — combining mine history, contamination data, cleanup status, community impact
- Fact-checks all numbers against source data, citations embedded
- Calculates “Shaft Score” — a composite metric of environmental recovery (water quality improvement + hazard reduction + community benefit)
- Generate:
- Interactive Mapbox/Leaflet maps with mine locations color-coded by status (hazard/in-progress/remediated)
- D3.js charts: water quality trends over time, funding allocation by state, cleanup progress bars
- AI-generated hero illustrations: cross-section diagrams of mine sites, pollution pathway infographics, remediation method explainers
- Before/after satellite comparison sliders for remediated sites
- State-level “report cards” with summary statistics
- Publish: Astro static site build → deploy to Cloudflare Pages (fast global CDN, free tier). Incremental builds for daily content, full rebuild weekly.
Tech Stack
- Static site: TypeScript + Astro (fast builds, component islands for interactive maps/charts)
- Maps: Mapbox GL JS or Leaflet + OpenStreetMap tiles (free tier sufficient)
- Charts: D3.js for custom data visualizations, Observable Plot for quick charts
- Image generation: DALL-E 3 or Stable Diffusion for hero illustrations; Puppeteer for chart screenshots (social sharing); sharp for image processing
- Data collection: Node.js scripts (fetch + cheerio for scraping, native fetch for APIs)
- Data storage: JSON files in repo (mine database), SQLite for local querying during build
- CI/CD: GitHub Actions (cron-triggered daily, webhook-triggered for breaking updates)
- Hosting: Cloudflare Pages (free tier: unlimited requests, 500 builds/month)
- Search: Pagefind (static search index, zero-cost)
- Newsletter: Buttondown (free tier: 100 subscribers) or Listmonk (self-hosted)
Monetization Model
-
Donations/Tips:
- Buy Me a Coffee — “Help us keep tracking the cleanup”
- GitHub Sponsors — for the open-source data pipeline
- Ko-fi — one-time and monthly supporters
- Projected: $200-500/month by month 3 (environmental/accountability projects attract passionate donors)
-
Newsletter Premium Tier:
- Free: Weekly digest of site updates, one featured story
- Premium ($5/mo): Full data exports, early access to state report cards, detailed environmental justice analysis
- Projected: 50 premium subscribers by month 6 = $250/mo
-
Data Licensing/Consulting:
- Sell cleaned, cross-referenced mine site database to environmental consultants, journalists, researchers
- API access for programmatic queries ($29/mo)
- Projected: $500-1000/mo by month 6
-
Sponsorship:
- Environmental remediation companies (Stantec, Arcadis, RESPEC)
- Outdoor recreation brands (Patagonia, REI — they fund environmental stories)
- State tourism boards near remediated sites
- Projected: $500-2000/mo by month 6
-
Affiliate:
- Water testing kits for communities near mine sites (legitimate public health tool)
- Books on mining history, environmental science
- Projected: $100-300/mo
- Projected month-1 revenue: $50-200 (early donor support, social media sharing)
- Projected month-6 revenue: $1,500-4,000 (with SEO traction, newsletter growth, first sponsorship)
Channel Soul & Character
Name: Shaft Report — punchy, double-meaning (mine shafts + getting the shaft from polluters), memorable.
Mascot: A grizzled canary wearing a hard hat and carrying a clipboard. The “Mine Canary” — historically the early warning system for miners. Now it’s the early warning system for communities. Illustrated in a slightly weathered, vintage-science-poster style.
Voice: A sharp-tongued environmental data journalist who grew up in coal country. Equal parts frustrated and hopeful. Drops hard numbers like a prosecutor but tells stories like a campfire narrator. Not preachy — more “look at what the data actually says” than “you should be outraged.” Occasionally darkly funny about the absurdity of century-old pollution still flowing. Calls out bureaucratic nonsense but celebrates genuine wins.
Example voice: “The good news: Pennsylvania’s Bennett Branch runs clean for the first time since William McKinley was president. The bad news: there are 43,000 more sites waiting. At this rate, we’ll finish somewhere around the year 2847. But hey, the money’s flowing. Let’s track where it goes.”
Opinion/Stance: The channel believes in accountability and transparency. It’s pro-cleanup, pro-data, pro-community. It doesn’t trust self-reported progress — it checks the water quality numbers. It names names: which states are spending effectively, which are sitting on federal money, which contractors are delivering results.
Running traditions:
- “Shaft Score Sunday” — weekly ranking of states by cleanup progress efficiency
- “The Canary Awards” — monthly recognition of best (and worst) remediation projects
- “Drift Report” — weekly roundup of new funding announcements and project completions
- “Deep Cut” — monthly long-form investigation into a single mine site’s complete history
- “pH Check” — regular water quality data updates for monitored sites
Visual style: Dark background (mine tunnel aesthetic) with amber/gold accent colors. Data visualizations in teals and oranges (representing clean and contaminated water). Topographic map contour lines as subtle background textures. Monospace fonts for data callouts. Custom mine site cross-section illustrations. Strong use of interactive maps with satisfying hover states.
Design Thinking
Visual differentiation: Unlike the beige government PDFs and basic Google Maps pins that dominate this space, Shaft Report looks like a premium data journalism site. Think ProPublica meets National Geographic. Dark mode with rich data visualizations that glow against the background.
Information architecture:
- Homepage: Live map with all tracked sites, “Shaft Score” leaderboard by state, latest featured story
- Site pages: Individual mine site profiles with full history, data charts, satellite imagery, cleanup timeline
- State pages: Report cards with aggregate data, funding tracker, site list
- Stories: Long-form narratives combining data with human stories
- Data hub: Raw data downloads, methodology, API documentation
Mobile-first: Map interactions simplified with tap-to-expand cards. Stories optimized for vertical scroll with sticky chart headers. Data tables converted to card views on mobile.
Shareability: “Shaft Score” state rankings are designed for Twitter/social debate. Before/after satellite sliders are screenshot-bait. Individual site cards with key stats are shareable units.
Trust signals: Every data point links to its source. Methodology page explains scoring. “Last updated” timestamps on every page. Government data source logos displayed prominently.
Launch Complexity: 3/5 (Multiple API integrations and data cross-referencing required, but all sources are free and well-documented. Map visualization adds complexity. ~3-4 weeks to MVP.)
Content Quality Score: 5/5 (Real data from federal databases, genuine environmental recovery stories with measurable outcomes, cross-referenced water quality monitoring, satellite imagery comparisons. This is investigative data journalism, not aggregation.)
Automation Score: 4/5 (Daily data collection and site updates are fully automated. Weekly narrative generation needs AI with good prompts and fact-checking against source data. The “Deep Cut” monthly long-form may need light human editorial review initially, but can become fully automated as templates mature.)
Revenue Potential: 5/5 ($11.3B federal program creates massive sustained interest. Environmental accountability attracts passionate donors. Data licensing is a real revenue stream. Outdoor/environmental sponsors are eager. Newsletter premium tier has clear value proposition for researchers and journalists.)
Total: 17/20
Why This Will Work:
Psychology: People are viscerally drawn to transformation stories — the river that went from toxic orange to trout-filled. Before/after content is one of the highest-engagement formats on social media. The mine exploration YouTube community (millions of views) proves massive appetite for mine content — Shaft Report captures the “what happens AFTER the explorer leaves” angle nobody covers.
Market logic: The $11.3B federal investment guarantees 15 years of continuous news. Every state has abandoned mines and active remediation programs, creating 50 state-level audiences. Environmental journalists currently spend hours cross-referencing government databases — we do that work for them. The environmental consulting industry ($44B market) needs this data synthesized. And the 78 million Americans near Superfund sites have a personal stake.
Timing: USGS just published the first national inventory (March 2025). States are deploying BIL funds NOW (PA and NM in March 2026). Environmental data journalism is a growing category (The Markup, ProPublica). But nobody has built the dedicated, automated, beautiful tracker for abandoned mines. First-mover advantage is wide open.
Risk & Mitigation:
| Risk | Likelihood | Mitigation |
|---|---|---|
| Government APIs go down or change | Medium | Cache all data locally in repo; multiple backup sources for key datasets |
| AI-generated narratives contain errors | Medium | All numbers fact-checked against source data programmatically; water quality claims verified against WQP API; disclaimer + correction policy |
| Low initial traffic | High (month 1) | State-specific content targets local SEO; pitch to environmental journalists; post site-specific findings to relevant Reddit communities |
| Data cross-referencing complexity | Medium | Start with a curated set of 100 “showcase sites” with complete data, expand systematically |
| Satellite imagery processing at scale | Medium | Start with manual selection of 10 key before/after sites, automate with Sentinel Hub API over time |