Italian Dispatch Model

How the model optimises battery dispatch across Italian markets


Getting Started

The Italian dispatch model builds on the core dispatch model framework. For an overview of the dispatch model architecture and optimisation approach, see the Core Dispatch Model documentation.

For an overview of the available revenue streams and market structures specific to Italy, see the Revenue Stack page.

The Italian model trades in three steps

Each Italian forecast is optimised against the prices of one bidding zone, in three steps:

  1. Day-ahead: the day-ahead market (MGP)
  2. Intraday auction: the second intraday auction (MI-A2), traded around the day-ahead position
  3. Real time: continuous intraday (MI-XBID), the balancing market (MB) and automatic Frequency Restoration Reserve (aFRR) energy, solved in successive 2-hour blocks

Day-ahead, intraday auction and balancing prices are known across the optimisation window. Continuous intraday prices are known within each 2-hour block, and as a 2-hour rolling average beyond it.

Italian balancing activations are capped by what batteries deliver today

The balancing market (MB) and aFRR energy pay for activated energy, so revenue depends on how much energy the battery is activated for. In each quarter-hour, the model caps that energy at an activation cap, expressed as a share of the battery’s rated power:

  • Measured on real batteries: the caps come from data published by GME (Gestore dei Mercati Energetici), the Italian power exchange, on the balancing activations of batteries in each Italian zone from December 2025 to June 2026
  • Volume and capture: each cap reflects both how much energy Terna activates and how much of it a battery can capture at each activation
  • By zone, direction and hour: each zone has an upward and a downward cap for every hour of the day, set separately for MB and aFRR energy
  • Same activated volume every year: the activated volume stays at its December 2025 to June 2026 level in every forecast year, and is shared among a growing fleet
  • aFRR coverage: aFRR energy is modelled in the North, Centre-North, Centre-South, South and Sicily. Calabria and Sardinia have no modelled aFRR energy.

The table gives the average cap before saturation, as a share of rated power:

Zone MB upward MB downward aFRR upward aFRR downward
North 2.1% 3.9% 1.7% 1.4%
Centre-North 0.6% 4.1% 1.5% 2.8%
Centre-South 1.9% 5.5% 0.8% 1.1%
South 1.0% 7.0% 1.9% 0.9%
Calabria 0.3% 1.2% 0% 0%
Sicily 1.4% 1.2% 0.2% 0.2%
Sardinia 2.6% 4.6% 0% 0%

MB downward caps are larger than upward caps in every zone but Sicily.

The battery offers its headroom, and the cap sets the share activated

In Italian forecasts, the battery offers the headroom left around its traded position, up to its rated power. The activated energy is that offer multiplied by the activation share, which is the cap divided by the rated power. The table shows a 50 MW battery with an upward cap of 5 MW, which is an activation share of 10%:

Traded position Largest upward offer Upward activation
Discharging 20 MW 30 MW 3 MW
Idle (0 MW) 50 MW 5 MW
Charging 20 MW 50 MW, the rated power 5 MW

Two further rules apply to Italian balancing:

  • One direction per quarter-hour: the balancing market and aFRR energy activate in one direction only in each quarter-hour
  • Continuous intraday rebalances the battery: in forecasts, continuous intraday is capped by the zone’s observed trading depth, and cannot be traded against an activation in the same quarter-hour.

Italian balancing caps fall as the battery fleet grows

More batteries compete for the same activated volume, so the model scales each zone’s MB and aFRR caps down as batteries are added after 2026:

  • Existing competition: the batteries added since 2026 are weighed against the flexible units that already compete in the zone’s balancing market
  • New battery weight: each new battery counts at 55% of its rated power, in line with the capacity today’s batteries offer in the balancing market
  • MACSE in the South and Calabria: a new battery counts at 30% there. Most capacity contracted under MACSE, Terna’s long-term storage contracts, sits in these zones, and MACSE projects keep only 20% of their ancillary revenue, so they are assumed to compete less.

The table gives the cap as a share of its 2026 level:

Zone 2030 2035 2040 2050
North 83% 76% 71% 70%
Centre-North 62% 53% 47% 46%
Centre-South 64% 53% 48% 47%
South 41% 32% 28% 13%
Calabria 82% 75% 71% 56%
Sicily 78% 71% 65% 36%
Sardinia 81% 73% 54% 27%

The South falls furthest, despite the lower MACSE weight, because it adds the most batteries relative to the flexible units already competing there.

Sardinian day-ahead spreads are aligned with history for 2026 to 2028

For Sardinia, the fundamentals model projects day-ahead spreads in 2026 to 2028 of 1.5 to 1.9 times those of the North, against 1.2 to 1.4 times over 2024 to 2026. For these three years, the dispatch model narrows each day’s Sardinian day-ahead price shape towards that day’s average price. This brings the Central scenario’s ratio of Sardinian to Northern daily spreads, measured between the four highest-priced and four lowest-priced hours, back to its 2024 to 2026 average of 1.32. Intraday prices move by the same amount, and daily average prices and balancing prices do not change.

Italian scenarios

The scenario adjustments, including the Italian day-ahead calibration factor, are on the Scenarios page. The activation caps, the Sardinia adjustment and the market rules on this page are the same in all three scenarios.

Italian results report intraday and balancing revenue as totals

Column Description
Intraday Revenue Revenue from the second intraday auction (MI-A2) and continuous intraday trading
Ancillary Services Revenue Revenue from the balancing market (MB) and aFRR energy, upward and downward combined

Simplifications in the Italian model

  • Balancing prices known in advance: real batteries bid into MB and aFRR without knowing the price.
  • Terna’s balancing needs held flat: future changes in activated volumes are not forecast.
  • One direction per quarter-hour: real activations can switch direction within a quarter-hour, so this rule is conservative.

  • Revenue Stack – Revenue streams available to battery storage in Italy
  • Core Dispatch Model – Global dispatch model architecture and methodology
  • Battery Dispatch – Battery inputs, constraints and optimisation windows
  • Scenarios – Assumptions behind the Central, Low and High scenarios