Electricity Maps: An Open-Source ESG App's Business Model

Most people have no idea whether the electricity coming out of their wall socket right now is mostly coal, mostly wind, or mostly nuclear — the grid doesn't label itself. Electricity Maps tries to fix that with a live map of carbon intensity by country and region, and unlike most climate apps that ask you to trust a dashboard — the core of any ESG app business model — this one lets you read the actual code and data pipeline that produces the number.

That's the part worth checking rather than taking on faith: is the "open source" label real, or just a badge on an otherwise closed product? The electricitymaps-contrib repository is where that question gets answered — in license terms, commit history, and which repos got archived along the way, not in marketing copy. It's also a rare chance to actually see an ESG app business model from the inside, instead of reading about one in an impact report.

What the map actually shows

The core repo describes itself as "a collection of parsers to collect and standardize electricity data such as production, exchanges and price from across the globe," and its README states the goal directly: "This project aims to provide a free, open-source, and transparent way to collect electricity data." That data feeds the public Electricity Maps platform, which the README says includes "our flow-tracing algorithm, estimation models, forecast engine and much more" — the pieces that turn raw generation and cross-border flow numbers into a single carbon-intensity figure per region, updated close to real time.

The config/zones directory is the clearest proof this isn't a handful of headline countries: it holds hundreds of zone configuration files — the project's own zone_names.json lists 291 zone codes — from national grids like DE.yaml (Germany) down to sub-national ones like AU-NSW.yaml (New South Wales) and CA-ON.yaml (Ontario) — real, browsable granularity rather than a marketing claim about "global coverage."

Where the numbers come from

Anyone can check the sourcing rules in DATA_SOURCES.md instead of trusting a summary. Real-time generation data comes from region-specific operators — Argentina's Cammesa, Australia's OpenNEM, Great Britain's Elexon BMRS — pulled, per the project's own description, from "public, free, and official sources." Installed generation capacity is cross-checked against international references like IRENA for renewables and IAEA PRIS for nuclear. Day-ahead electricity prices come from market operators such as France's RTE and Japan's JEPX, and wind/solar forecasting inputs draw on the US National Weather Service's Global Forecast System. That's an unusually citable methodology for a consumer-facing climate number.

Volunteers write the parsers, not a closed team

Coverage of hundreds of grid zones isn't maintained by one internal team — it runs on outside contributions, and CONTRIBUTING.md lays out exactly how a new parser gets in: fork the repo, write the parser against the documented interface, validate it locally, then open a pull request. First-time contributors need CI approval from the Electricity Maps team, after which the change goes through both automated tests and manual review before a release ships it to production. It's a real, moderated pipeline rather than an open free-for-all — which is also why coverage gaps and stale parsers are a known, trackable category in the project's issue tracker rather than a hidden one.

The license is doing real work

Electricity Maps ships under the GNU Affero General Public License v3, not the more common plain GPL or MIT. That choice matters specifically for a hosted data product: ordinary GPL only forces you to share source code if you distribute the software, and running a modified version on a server you never hand to anyone doesn't count as distribution. The AGPL closes exactly that loophole — its own preamble says it "is designed specifically to ensure that, in such cases, the modified source code becomes available to the community." In practice, that means a competitor can't quietly fork the parser engine into a closed commercial API without publishing their changes back.

Consolidated from older projects, not a clean launch

The org's GitHub history reads like a project that consolidated rather than one that appeared fully formed. Two archived repositories tell that story in dates and descriptions rather than a "founding myth" retold in a blog post: co2signal-docs, documentation for a "CO2 Signal API," was archived on February 3, 2021, and bloom-contrib — described in its own README as "Making carbon footprint data available to everyone" — was archived on March 31, 2021. Both now live under the electricitymaps GitHub organization, which today describes itself simply as "the world's most comprehensive electricity data platform," based in Denmark. Read together, the archive dates show a rebrand and consolidation around 2021 rather than a single clean launch — worth knowing if you see "CO2 Signal" referenced in older writeups.

What's open, and what quietly isn't

Here's the honest split. The electricitymaps-contrib repo — the parser engine and data pipeline — is genuinely open under the AGPL, with public issues, public pull requests, and a public review process. The consumer-facing map and mobile apps are a separate story: a now-archived frontend project, electricitymaps-contrib-rewrite, was folded back into the main product in January 2023, and its own README pointed users to "app.electricitymaps.com, or download the app on Google Play or App store" — but that frontend code no longer lives in the open-source repo today.

The most recent tagged release, v1.322.0 on April 28, 2025, explicitly folds in the removal of already-deprecated mobile-app and web-frontend code from the open-source tree. In other words: the data engine that computes the carbon-intensity numbers is open and auditable; the app you'd actually install is not built from that same public repo.

That's a common and defensible open-core structure, but it's not "everything about this app is open source." The project's own commit history is what makes that distinction checkable at all, rather than a matter of taking their word for it.

Still actively maintained

An "open" climate data project is only worth trusting if it's still being run. The commit history on master shows a commit as recent as September 1, 2026 — three days before this was checked — with dense activity through all of August 2026 before that, plus 113 open issues and 11 open pull requests actively being worked. The repo itself carries roughly 4,000 stars and 1,100 forks.

That's a live, worked-on project, even though its formal version-tagged releases have slowed since early 2025 — a gap worth noting honestly rather than glossing over. A slower release cadence combined with continuous unversioned commits is a different, less legible signal than a project that ships regular numbered releases.

What this ESG app business model generalizes to

Electricity Maps is a useful test case for "ESG app business model" claims generally: instead of a glossy impact report, its main credibility artifacts are a license file, a commit graph, and a directory of hundreds of zone configs anyone can open. That's the same discipline worth applying to NVDA's screen reader or any other "for good" tool that calls itself open source — check what's actually in the public repo, check what got quietly archived or pulled out, and don't assume the free consumer app and the open-source engine underneath it are the same thing just because they share a name. The mechanism generalizes past climate data, too — the same "read the license, read the commit graph, read what got archived" checklist is what separates a gamified habit app that's still alive from one running on marketing inertia alone.

If you're evaluating any climate or ESG tool that advertises itself as open source, that's the concrete step: go to its repo, find the license file, and check whether the thing you'd actually use is built from the code you can actually read.