Inspiration

Last year California threw away 3.77 billion kWh of clean energy, enough to run about 620,000 homes for a year. Roughly 80% of it was cut because power lines were full, not because the sun stopped shining. Across the US, grid bottlenecks cost about $12 billion in 2024, and that cost lands on everyone's electric bill.

The surprising part: every power line has one fixed limit, set for the worst day of the year (about 100 °F, full sun, no wind). Most clean energy is wasted on cool spring days, when that same wire could safely carry much more. We wanted a map that shows this to anyone, not just utility engineers.

What it does

Click anywhere within 100 miles of Westminster, CA and Mapergy scans the circle:

  • What's there: every power plant, power line and substation, from public federal and open data.
  • How much more the lines can carry right now: live weather plus the IEEE 738 line-heating model gives a real-time rating. For example, "1.31×" means about 31% more than the fixed limit.
  • What's happening live: weather, CAISO and EIA grid demand, wildfires and earthquakes, cross-checked between independent sources, plus live local grid news.
  • What to do about it: a step-by-step optimization workflow with fixed, written rules (weather-based rating, then batteries, then reconductoring, with a new line as the last resort) that ends in a ranked shortlist.
  • Report a problem: pick what's wrong (for example "tree touching a wire") and Mapergy writes a formal letter, finds the right utility for that spot, and opens an email already addressed to it, ready to send.
  • For infrastructure owners: sign in, claim an asset, and track it through 10 lifecycle gates.

Every number traces back to the exact source file it came from. If we don't know something, the app says "unknown" instead of guessing.

How we built it

  • Frontend: Next.js 15, React, TypeScript, Tailwind, Leaflet (2D) and MapLibre GL (3D terrain).
  • Data: EIA-860 plants, HIFLD and OpenStreetMap lines and substations, NWS and Open-Meteo weather, CAISO and EIA grid data, NIFC wildfires, USGS earthquakes, FEMA risk index, and Google News / GDELT for local news.
  • Physics: the IEEE 738 heat balance. Heat in equals heat out:

$$I^2R(T_c) + q_s = q_c + q_r \quad\Rightarrow\quad I = \sqrt{\frac{q_c + q_r - q_s}{R(T_c)}}$$

Solving for the current \( I \) gives the line's rating for the real weather instead of the worst-case day.

  • Evidence pipeline: a Python/TypeScript ETL saves every download as a SHA-256 snapshot. Each claim in a report points to its exact source, and any report can be rebuilt offline from the stored data to prove nothing changed (deterministic replay).
  • Backend: Supabase (Postgres with row-level security, plus storage) for datasets, weather history and owner claims; Clerk for sign-in; Netlify for hosting.
  • Testing: 165 app tests and 350+ ETL tests, including a check that our TypeScript line rating matches the Python model to within 0.06%.

Challenges we ran into

  • Messy public data: HIFLD and OpenStreetMap disagree on substations, so matches need review instead of being trusted blindly.
  • Live APIs fail: EIA timed out on the shared demo key, GDELT rate-limits to one request every 5 seconds, and search results drift to old or off-topic stories. We added fallbacks, caching and relevance filters (for example, "Long Beach Peninsula" is in Washington state, not California).
  • Finding real contacts: many utilities publish no email address. We used only verified addresses and route SCE, SDG&E and PG&E reports to the CPUC Consumer Affairs Branch when the utility lists none.
  • Silent fallbacks: one bug made the site quietly read local files instead of Supabase. Now we test that data really comes from the source we claim.

Accomplishments that we're proud of

  • A working live map that turns a hard engineering formula into one number anyone understands.
  • Reports that go straight to the right people, written like a real formal complaint.
  • Evidence you can verify: every figure links back to its public source.

What we learned

  • The grid's biggest limit is often temperature, not wire capacity, and weather changes everything.
  • Being honest about unknowns builds more trust than a confident guess.
  • Real-world data integration is most of the work.

What's next for Mapergy

  • Expand beyond Southern California.
  • Partner with a utility to compare our weather-based ratings with real line sensors.
  • Push alerts when a nearby line is congested or at wildfire risk.

Built With

Share this project:

Updates

Submission history