Forecasting prices: short-term drivers and long-term saturation
Ancillary services provide a promising revenue opportunity for BESS operators in Germany and Iberia. However, prices in these markets are forecast to decline as battery buildout grows, as has been observed in markets with greater BESS buildout like GB and ERCOT.
We forecast ancillary services prices for all applicable markets in Germany, Spain, Portugal, and GB. Because short-term price patterns and long-term saturation behave differently, we forecast these prices in two steps.
Step 1: Predicting baseline prices (short-term dynamics)
Ancillary services prices are driven by the demand for reserves, the technologies providing them (especially batteries), and the market rules.
We train a regression model on recent historical data, using features such as time of day, seasonality, and wholesale prices. We then apply this model to our forecast wholesale prices to predict baseline ancillary-services prices.
Step 2: Adjusting for saturation (long-term effects)
We expect the prices of capacity-based ancillary services markets to fall as these markets become ‘saturated’, as observed in other markets such as GB and ERCOT. A power-law model is observed in these markets, which we apply to the forecast prices produced in step 1. The power-law model predicts that as battery buildout in a country exceeds the size of the ancillary services market, the prices in that market fall.
- Smaller markets (e.g., FCR) prices fall earlier because they saturate sooner
- We do not expect prices to fall all the way to zero—they will stabilise at a floor related to cycling and degradation costs
- We do not expect prices in energy-based markets (like aFRR Energy) to saturate since these markets require actual energy delivery, so will remain tied to the cost of supplying energy
Note: We will refresh the model regularly as new market data and battery build-out information become available.