GridSim methodology, in plain terms

1. GridSim estimates what a battery would earn at any point on the Texas grid, using only public market data, and checks that estimate against real batteries it never saw.
2. Batteries earn by charging when power is cheap and selling it back when it's expensive. Over half the US battery capacity planned for 2026 is in Texas.
3. Most of Texas runs on its own grid, ERCOT. It changed how batteries are scheduled in December 2025, so GridSim uses only the months since.
4. Power is priced separately at each point where a plant or battery connects. So what a battery earns depends on where it plugs in.
5. With hindsight, work out the most a 100 MW, two-hour battery at one node could have earned if it had known every price in advance.
6. Real batteries don't know tomorrow's prices. They captured 18% to 48% of their ceiling, so GridSim scales the ceiling down to match.
7. Tested on 315 real batteries the method never saw, every one in ERCOT since the December 2025 change. On batteries of 20 MW or more the median miss was 23% and 43% landed within 20%, with no lean either way: right on average, wide for any one site.
8. Most of what separates batteries is how they're run, not where: the operator explains about 72% of the spread, the site about 13%. GridSim screens the site. It ranks sites with a rank correlation of 0.56, and when it puts one 30% above another it's right 80% of the time.
9. More batteries chasing the same hours means less for each. That forecast failed its own July test, so it shows a direction, not a number.
10. GridSim is a screening estimate, not financial or investment advice. For a report on a node you're looking at, reach out: contactmaldrich@gmail.com