Energy by Cogen dispatches 1.42 TWh/year across a fleet of 20 plants (245 MW registered) at a weighted-average price of 72.4 €/MWh versus an arithmetic spot mean of 65.0 €/MWh: a positive capture spread of +7.43 €/MWh, equivalent to ≈10.5 M€/year of additional revenue above the flat baseline. They rank #2 in the MIBEL cogen peer group for 2025 — a top-tier operator in their segment.
Reopening the dossier since our last collaboration, the analysis across 17,354 hours of I90 data identifies five residual levers with an aggregate impact of 1.2 to 3.0 M€/year.
Reopening the Cogen dossier after our last collaboration. The data confirms the fleet remains #2 in capture spread among MIBEL cogen operators in 2025 (only surpassed by a marginal-size cooperative). 2025 dispatch: 1.42 TWh at +7.43 €/MWh above the flat baseline. The price-dispatch curve is what you would expect from a sophisticated bidding desk: 138 MW during negative-price hours (heat-bound) and 217 MW during «very_high» hours. ρ = +0.449 is exceptional for a thermal cogen operator — confirming the desk's bidding goes beyond merely serving the heat customer's demand.
The analysis identifies three operational levers and two regulatory-defence services with an aggregate impact of 1.2 to 3.0 M€/year over 2024-2026 operations.
The data confirms what we already knew: Cogen runs the fleet as a thermal aggregator, not a must-run. The question is how much of the residual depends on the heat customer and how much can be captured through finer bidding, secondary reserves, and more dynamic gas/biomass mix.
The 138 MW floor during give-away hours is the «heat floor». When the spot is negative, Cogen still dispatches 138 MW because the heat customer needs steam. That is ≈78 MW of marginal generation at negative cost (138 MW total minus 60 MW that could technically be stopped). Over 782 hours of spot < 0 at an average price of −1.56 €/MWh, that is −95 k€/year of lost revenue — but it is also −95 k€/year if we choose not to stop. The question is not whether to stop: it is whether there is an alternative heat host or thermal battery to avoid burning gas in those moments.
The extreme peak flattens — a sign of a ramping ceiling. During «very_high» hours (162 €/MWh) the fleet delivers 217 MW; during «extreme» (212 €/MWh) only 211 MW. That flattening suggests some plants do not reach those punctual peaks in time. Rescheduling the maintenance calendar (lever #4) could capture that surplus.
Five levers. The first three are operational (better bidding and reserve markets); the last two are regulatory defence (capacity and fuel mix). As an existing customer, levers 1 and 4 are the most aligned with prior conversations.
The current capture spread of +7.43 €/MWh is excellent, but the P10–P90 range within each price bucket is ~50 MW — indicating dispersion of behaviour across plants. A centralised optimisation taking each heat-host's thermal forecast + the spot forecast + technical ramping constraints can shave 0.5 €/MWh of recoverable sub-optimal spread.
Over 1.42 TWh/year: 0.7 to 1.5 M€/year additional capture.
Daily MILP per plant taking: thermal forecast for each heat-host (process industry, district heating), 15-min spot forecast, technical constraints (ramping, technical minima), gas price + EUA. Output: optimal bid curve per plant and time-of-day. Intraday re-optimisation.
The fleet's large plants (TAE2COG 32 MW, TOLEXTR 16.6 MW, TVALPBA 16.5 MW, etc.) have an ideal technical profile for secondary reserve — synchronous, centrally controlled, with ramping headroom. If only 30 % of the fleet participates today, scaling to 70 % means ≈40 MW additional.
2025 secondary prices: upward ~65 €/MW·h, downward ~22 €/MW·h. Scenario: 40 MW × 4 000 h × 70 % of theoretical price: 0.3 to 0.8 M€/year.
Centralised system evaluating daily which plants can offer up/down reserve given their thermal program. Settles per-plant and per-heat-customer (revenue-share). REE compliance.
Cogen has 2 plants flagged as Biomass (REGBRE 8 MW, REXTRA 9 MW) and 18 Natural Gas plants. If any of the gas plants have biomass co-firing capability (common in Cogen IPP), a daily advisor that decides the optimal mix based on the TTF + EUA + solid-biomass spread can capture 0.5–1.5 €/MWh on the eligible plants.
Assuming 4 plants with dual capability and 100 GWh/year between them: 0.1 to 0.3 M€/year. Small in economics, high in ESG positioning.
Model combining: TTF spot + forward price, EUA price, per-plant biomass logistics cost, biomass supply calendar. Output: recommended co-firing % per plant for D+1.
The flattening of the «extreme» bucket (217 MW at very_high → 211 MW at extreme) suggests outages or technical limitations clustered at punctual moments. Rescheduling the fleet's annual maintenance plan to move outages from expensive months (Jan-Feb-Jul-Aug) to cheap months (Mar-Apr-May) captures the monthly differential.
Average monthly spread between the top-2 and bottom-2 months in 2025: ≈40 €/MWh. Moving 50 GWh/year of maintenance from expensive to cheap months × 40 €/MWh × 30 % capture: 0.15 to 0.30 M€/year.
Annual planner with per-plant technical windows + spot forward curves. Allows simulating re-scheduling scenarios and seeing economic impact. Integrates with existing CMMS.
The new capacity mechanism for thermal plants has strict availability and response requirements. Cogen, with 245 MW of firm capacity, is a natural candidate — but it requires documenting historical availability + REE tests. A structured audit can ensure no plant is left out due to paperwork issues.
Expected capacity payment: 30–60 k€/MW·year × eligible MW. If Cogen qualifies 200 of the 245 MW: 50 to 120 k€/year of incremental capacity payment.
Continuous per-plant availability tracking vs mechanism requirements. Early non-compliance alerts. Automated REE reporting. Auditor version for periodic review.
Cogen is the dominant operator in the MIBEL cogen segment — the next ranked retailer-cogen (Cogeneración de Beguda) has 36 GWh/year vs Cogen's 1.42 TWh. That market-share concentration is positive (REE negotiation power, scale to invest in optimisation) but also carries a risk: any regulatory change in the segment (capacity mechanism, CHP charges, EUAs) hits directly and without diversification. Levers #2 and #5 exploit the scale; #1 hedges that regulatory risk by turning Cogen into a market operator, not just a heat-host operator.
| # | Service | Data maturity | Complexity | Time-to-value | Impact €/year |
|---|---|---|---|---|---|
| 01 | Heat-Aware Bid Optimiser — per-plant MILP with thermal constraints — trading desk | Medium (plant telemetry + heat-host SCADA, requires coordination with industrials) | High | 8–12 months | 0.7 – 1.5 M€ |
| 02 | Floating-reserve platform for cogen fleets — grid services | Medium-high | Medium | 5–8 months | 0.3 – 0.8 M€ |
| 03 | Fuel Mix Daily Advisor — gas vs biomass per plant — operations | High (gas consumption + biomass billing) | Medium | 3–5 months | 0.1 – 0.3 M€ |
| 04 | Maintenance Window Optimiser + What-If Simulator — planning | Medium (maintenance log) | Low-medium | 2–4 months | 0.1 – 0.3 M€ |
| 05 | Capacity-Mechanism Compliance Dashboard — regulatory | High | Low (technical) / Medium (regulatory) | 2–3 months | 0.1 – 0.1 M€ |
| Aggregate impact (adjusted for cannibalisation) | 1.2 – 3.0 M€ | ||||
The recommended order follows a defence-vs-offence logic: levers 5 and 2 first (capacity audit + extend reserve — both are defensive revenue, fast); then lever 4 (maintenance optimiser — short implementation); then levers 1 and 3 (higher impact but long implementation). Cannibalisation: low across the five (≈15 %); 1 and 2 share some plant capacity but the revenues are additive.