Loading…
A look at California's grid connection queue
What happens to power projects waiting to join California's grid
Before a new solar farm, wind farm or battery can send power to the grid, it applies to connect and waits in a while the grid operator studies it. Headway looks at what happened to the … projects that applied under the old rules, and uses that history to estimate, at each , how likely a new project is to be built and how long it tends to take.
Loading the history…
The history
Most projects leave the queue before they're built
Each dot below is one project that applied to connect to California's grid, placed by the year it applied. It plays through what happened to them; pause or step through at any point.
Location
How crowded the queue is varies a lot by substation
Each bar below is a substation. Its height is the of projects realistically waiting there: each waiting project counted by its chance of being built, given how long it has already waited.
Working out every substation. This takes a few seconds the first time.
Method
How the numbers are made
Traceable
Each number links to how it was worked out and the exact rows of the grid operator's data it came from.
Checked by code
A language model drafts each written summary. Code then recalculates every number in it from the data, and anything that doesn't match is left out.
Reviewed by a person
Interpretations are labelled as opinions and wait for a reviewer to accept, reject or reword them. Leaving a project out recalculates everything that depends on it.
About this project
What it is
An independent project exploring how a language model can write analysis whose every number is checked against public data. It isn't affiliated with the grid operator.
Data
CAISO's interconnection queue report. Substation positions from OpenStreetMap and public filings; county boundaries from the US Census Bureau.
Method
For each group of similar past projects, the share built by each year after applying. Projects still waiting count only for the years they've been watched. Likely ranges come from resampling, and past predictions are checked against what happened.
Limits
The history covers projects that applied before the ; newer projects can't be scored yet. It describes past outcomes, not the physics of the grid, and doesn't replace a power-flow study.