Skip to content
Projects / Phase 4.3 · JD to Projects

Grid Battery Valuation, Dispatch and Hedging

What is the battery worth, how should it be dispatched and hedged, should it be tolled?

Open the projectSee all projects

Headline result

Dispatch, hedging and valuation of a 100 MW / 200 MWh battery at ERCOT’s North hub, 2019 to 2026. Its ten-year value fell from a $126.1M peak (Sep 2023) to $13.6M as ancillary prices collapsed. Degradation-aware dispatch cut cycling 63%, giving up $0.5M of revenue to avoid $3.45M of wear.

Figures

  • Monthly median daily capture ratio of four dispatch policies against perfect foresight, 2019 to 2026.
    Capture fell from about 0.95 in development to 0.65 in evaluation; dispatch that ignores degradation falls furthest after ERCOT’s RTC+B market redesign.
  • Monthly battery revenue by line: ancillary services, energy, real-time deviation and degradation, symmetric log scale, 2019 to 2026.
    Ancillary services were nearly all revenue before 2023 and shrank after; merchant revenue fell from $344/kW in 2023 to $21/kW in Jan–Sep 2026.
  • Tornado of ten-year value drivers at end-2022, and quarterly valuation updates from 2023 to 2026.
    Ancillary-service saturation dwarfs every other driver; the ten-year value rose from $82.9M to $126.1M, then fell to $13.6M by Sep 2026.
  • Monthly mean day-ahead ancillary-service prices by product, log scale, with dated ERCOT rule changes, 2017 to 2026.
    Regulation-up and responsive reserve averaged about $18–21/MW-h in 2019 to 2022 outside Winter Storm Uri, and $1.5–2.2 after the RTC+B redesign.

Data used

  • ERCOT HB_NORTH day-ahead, real-time and ancillary-service prices via Grid Status, Jun 2016 to Oct 2026
  • EIA-930 ERCOT load, wind and solar, Jul 2018 to Oct 2026
  • Henry Hub daily spot via FRED and EIA, 1997 to Oct 2026

Built from Energy Quantitative Analyst job descriptions

The role asks for

  • Energy asset valuation
  • Dispatch and bidding optimisation
  • Power price forecasting
  • Hedging, tolling and revenue risk
  • Python, SQL and statistics
  • Clear communication with commercial teams

This project uses

  • LEAR day-ahead price forecasting
  • LP dispatch with degradation cost
  • Stochastic planning with a CVaR penalty
  • Least-squares Monte Carlo valuation
  • Revenue-at-risk backtesting
  • Toll and revenue-floor pricing

Members open the full write-up, notebook, tested code and data in the project reader.