How single-asset battery and solar revenues are modeled in SPP
The dispatch model produces site-specific revenue forecasts for individual battery and solar assets, using day-ahead zonal prices from the production cost model. It co-optimizes energy arbitrage against operating reserve revenues from the perspective of a single asset owner. The forecast runs on the day-ahead market only: there is no real-time settlement, so real-time price volatility is not captured.
It is the same single-asset dispatch optimization used across Modo’s market regions, applied to SPP prices. See the core Dispatch Model for the shared methodology.
Batteries, solar or both can be modeled at one site
A site can hold a battery, a solar array, or both behind one grid connection:
- Standalone battery: energy arbitrage and operating reserves.
- Standalone solar: energy sales, plus any contracts on the generation.
- Co-located solar and battery: the two share the grid connection. On an AC-coupled site the battery behaves much like a standalone one. On a DC-coupled site it can also store solar output the inverter would otherwise clip.
Solar output is modelled from ERA5 irradiance at the site, for the weather year at the chosen yield exceedance level (P50 by default). The profile is then reshuffled onto 2023, the weather year behind the SPP price forecast, so solar generation lines up with the prices it sells into. Solar is available in forecasts only: backtests model the battery alone. See Solar and Co-located Solar for the shared methodology.
Revenue streams
| Stream | Method | Source |
|---|---|---|
| Energy arbitrage | Co-optimized | Day-ahead zonal LMPs from the production cost model |
| Operating reserves | Co-optimized | Predicted day-ahead reserve-zone prices |
| Solar generation | Co-optimized with the battery on co-located sites | Day-ahead zonal LMPs, adjusted by any contracts |
| Contracts | Settled on the asset | Terms set for each run |
SPP has no centralized capacity market, so there is no capacity-market revenue stream. A storage asset’s contribution to resource adequacy is reflected through capacity accreditation in the Capacity Expansion Model, not as a market payment.
Energy arbitrage
The battery charges and discharges against its day-ahead zonal LMP, subject to state-of-charge, power, round-trip efficiency, and cycling constraints.
Forecasts build imperfect price foresight into the optimization. The battery schedules against the energy price smoothed over a centered three-hour window, so it sees the shape of each day but not the exact hour of each peak. The schedule is then settled at the actual, unsmoothed prices. Backtests schedule against the actual prices directly.
Operating reserves
Four SPP ancillary-service products are modeled. Regulation is procured system-wide, while the contingency reserves are defined across SPP’s reserve zones:
| Product | Procurement level |
|---|---|
| Regulation up | System-wide |
| Regulation down | System-wide |
| Spinning reserve | Reserve zone |
| Supplemental reserve | Reserve zone |
Reserve prices are not co-optimized in the production cost solve. They are predicted statistically (a gradient-boosted model on historical SPP day-ahead ancillary-service clearing prices), driven by the energy price in the relevant reserve zone. They then decline as the battery fleet grows and competes for a finite reserve requirement. The decline starts after July 2026, the last month of observed prices, which show no saturation yet.
SPP also operates ramp-capability and uncertainty products, but these are not valued in the forecast. Only regulation, spinning, and supplemental reserves earn revenue alongside energy.
Reserve awards are capped at the battery’s share of the requirement
SPP procures a fixed volume of each reserve product, and its battery fleet is set to grow well past it. A single battery cannot sell its full capacity into these markets every hour.
The model therefore caps each reserve award at an acceptance rate: the share of an offer the market can absorb. The rate is the requirement batteries can serve divided by the battery capacity competing for it, capped at 100%.
- Requirement: a fixed SPP-wide volume for each product, held flat across the horizon. SPP does not publish reserve awards by resource type, so batteries are assumed able to win the whole requirement.
- Competing fleet: existing batteries, committed queue projects and the batteries the capacity expansion model builds, each from its commercial operation month. Each counts at half its capacity, since not every battery offers reserves.
- One rate across SPP: there are no zonal requirement figures, so every reserve zone carries the same rate.
The rate falls quickly as committed storage comes online, then flattens once the capacity expansion model’s build slows. Contingency reserves have the larger requirement, so more of each offer is accepted than for regulation.
The reserve price decline and the acceptance rate are driven by the same competing fleet, so a battery’s reserve revenue per megawatt falls through both price and volume as the fleet grows. The cap applies to forecasts only. Backtests settle against historical prices with no cap, because there are no published battery awards to set one from.
Contracts
Solar can carry contracts that change what it earns from the wholesale market: merchant, fixed payment per megawatt-hour, the Production Tax Credit (PTC), renewable energy certificates (RECs), and power purchase agreements (PPAs) at a fixed price, with a floor, or with a cap and floor.
See Revenue Contracts for how each contract settles.
Calibration
A single calibration factor of 80% is applied to every modeled battery revenue stream. Solar generation revenue is not scaled. The factor brings the optimizer’s result down to the revenue a real, actively traded asset is expected to capture in practice.
SPP has little realized battery operating history to calibrate against directly, so the factor is carried over from Modo’s benchmarking in more established storage markets, principally Great Britain and Australia’s NEM, where years of modeled and actual asset revenues can be compared. It will be revisited as SPP storage history accumulates.
Outputs
Monthly and annual per-asset metrics:
- Revenue by stream, including contracts
- Cycling: charge/discharge cycles and degradation
- Solar: generation, capture price and clipping losses
The detailed results also report, for each period, the cleared price of every stream, the day-ahead energy price and each ancillary-service price, alongside the dispatched charge and discharge volumes and state of charge.