Integrated model: PV orientation + BESS + day/night split monthly sufficiency + grid quality + system cost. Reference year: 2025 REE actuals (Panel 3). Model represents 2025 scenario: end-2025 installed capacity (48 GW solar, 33 GW wind), 256 TWh demand.
By Isgar Bos · isgarbos.com
This is a directional tool, not a precision tool. Move the sliders to discover how policy levers interact — solar orientation, BESS, grid corridor, electrification, weather years. The structural relationships (solar capture rate falls with scale; wind displaces gas more than solar; corridor unlocks curtailed renewables; demand growth rescues solar's economics through the S-curve) are robust and reflect Spanish grid physics. The absolute numbers in €/tCO₂, €/MWh and TWh are indicative — calibrated to REE 2025 and Aurora/Ember/IRENA published values, but not predictive at single-decimal precision. Use this to understand what matters and why; for project-grade investment decisions, refer to a full hourly-dispatch model (PLEXOS, Aurora) with stochastic weather years.
| Source | TWh | % mix | Installed | Note |
|---|---|---|---|---|
| Wind | 58.8 | 21.6% | 33 GW | REE confirmed. Post-blackout 4 TWh curtailed. Normal year = 58.8 + 4 = 62.8 TWh. Model uses 62.8 TWh. |
| Nuclear | 51.7 | 19.0% | 7.1 GW | REE confirmed. Flat baseload; phase-out 2027–2035 (contested — Almaraz extension requested Oct 2025) |
| Solar PV (grid) | 50.2 | 18.4% | 48 GW | REE confirmed. Post-blackout 3 TWh curtailed. Normal year = 50.2 + 3 = 53.2 TWh. Excl. ~11 TWh self-consumption. Model uses 53.2 TWh. |
| Gas CCGT (peninsular) | 45.7 | 16.8% | 26 GW | REE confirmed. Post-blackout: REE curtailed 7 TWh wind+solar and ran 7 TWh extra gas to compensate. Normal year = 45.7 − 7 = 38.7 TWh. |
| Hydro (conventional) | 27.8 | 10.2% | 12 GW | REE confirmed. Essentially normal year (avg ~28 TWh). Model slider set to 28 TWh. |
| CHP / cogen | 16.3 | 6.0% | — | Derived (6% × 271.9). Industrial byproduct — zero marginal electricity cost; runs regardless of grid conditions. |
| Pumped hydro (output) | 5.9 | 2.2% | 8 GW | REE confirmed: 5,886 GWh discharged, 9,204 GWh charged. Net consumer of electricity (−3.3 TWh). Modelled separately. |
| Island fossil (gas/oil) | 8.0 | 2.9% | — | Canarias + Baleares + Ceuta/Melilla. NOT in peninsular CCGT figure. Not modelled — island systems are isolated. |
| Solar thermal (CSP) | 2.4 | 0.9% | 2.3 GW | Derived (0.9% × 271.9). Concentrated solar — dispatchable with thermal storage. |
| Biomass + biogas | 4.1 | 1.5% | — | Derived. Stable baseload from organic waste and agricultural residues. |
| Coal | 1.1 | 0.4% | <1 GW | Derived. Historic low — 50% reduction vs 2024. Near phase-out. |
| Total metered generation | 271.9 | 100% | 142 GW | REE official figure. Excl. self-consumption (~11 TWh estimated, behind-meter). CO₂ emissions: 29.5 Mt (REE confirmed). |
| Pumped storage consumption | −9.2 | — | — | REE confirmed: 9,204 GWh consumed to pump water uphill. Counted on demand side. |
| Net exports | −13.0 | — | — | REE confirmed: 12,794 GWh. 4th consecutive year net exporter. Not attributable to any single source. |
| Transmission losses | ~2.7 | — | — | Estimated ~1% of generation. |
| Domestic demand (metered) | 256 | — | — | REE confirmed: 256,086 GWh. Growing 2.8% vs 2024. Balance: 271.9 − 9.2 − 13.0 − 2.7 ≈ 247 TWh delivered + 9 TWh rounding/other. |
| Structural curtailment | ~2 | — | — | Normal grid congestion. Aurora 2026 estimate: ~3 TWh/yr structural going forward. |
| Post-blackout curtailment | ~7 | — | — | One-off 2025: REE curtailed renewables to keep gas running for grid inertia post April 28. July: 11% curtailment rate. Returns to ~3 TWh structural in 2026. |
|
Model normalisation: Wind/solar set to normalised 2025 (62.8/53.2 TWh) — post-blackout 7 TWh curtailment restored to RE generation. Other firm sources (CSP 2.4 + biomass 4.1 + coal 1.1 = 7.6 TWh) added to supply pool. Net exports default 13 TWh (REE actual). At defaults: gas ≈ 45 TWh, matching REE metered 45.7 — i.e. model reproduces the actual blackout-distorted year. Set exports = 0 to see "closed system" gas of ~32 TWh; set wind/solar back to metered 58.8/50.2 to model the actual 2025 dispatch. | ||||
| Source | Capex €M/GW | LCOE €/MWh | CF % | Note |
|---|---|---|---|---|
| Solar PV | €700M/GW | €32/MWh | 19% | Lowest in Europe. Capture rate ~55% at 48 GW — falling with scale |
| Onshore wind | €1,200M/GW | €50/MWh | 28% | Best sites Spain. EU avg auction €76/MWh. Seasonally complementary to solar |
| Nuclear life extension | €500M/GW | €33/MWh | 83% | Cheapest firm capacity. 30yr life extension. Phase-out 2027–2035 contested |
| Nuclear new build | €6,000–8,000M/GW | €90–120/MWh | 85% | European new build 2025. Not realistic for Spain this decade |
| Conventional hydro | — | ~€15/MWh | variable | Fully developed. Existing fleet amortised. No new sites available |
| Pumped hydro (new) | €1,500–2,000M/GW | €50–80/MWh | seasonal | Closed-loop only. Long life, seasonal storage, provides inertia |
| Field BESS | €150/kWh | €65/MWh | daily | Co-located at PV site. Daily cycling. No grid permit needed |
| Grid corridor (S→N) | €250–500M/GW | — | — | Payback 3–7 yrs on gas savings alone. Permitting is the bottleneck |
Every number below is a modeller choice or a literature-sourced constant. Change any of them and policy ranking can flip. Listed for transparency.
| Coefficient | Value | Used for | Source / Justification |
|---|---|---|---|
| Solar gas-displacement rate (base) | 15% at 256 TWh demand | Panel 8 €/tCO₂ for solar | Aurora 2025 anchor. S-curve in demand (cumulative-normal, inflection 320 TWh, peak +40pp at 400 TWh). Reflects: noon already saturated at low demand → flat slope; mid-demand electrification absorbs noon → steepest gain; very high demand re-saturates → tapers. Corridor coupling: zero at default sliders (gas barely runs at midday today), grows with joint stress of demand × solar build, capped +8pp. Captures that S→N corridor matters for solar gas-displacement only when both electrification and solar expansion stress the transmission. |
| Wind gas-displacement rate (base) | 85% at 256 TWh demand | Panel 8 €/tCO₂ for wind | Aurora/Ember anchor. S-curve, peak +10pp at 400 TWh. Wind already high — small headroom. |
| Solar capture rate (base) | 52% at 48 GW, 256 TWh, pure south | Effective solar cost | Aurora 2025 anchor. S-curve in demand (peak +30pp). Linear penalty for scale (−0.3pp/GW above 48). Linear bonuses for E/W (+15pp), BESS (+10pp cap), pumped (+5pp), corridor (+5pp). |
| Wind capture rate (base) | 85% at 256 TWh | Effective wind cost | EU avg 2024. S-curve, peak +5pp at 400 TWh. |
| Grid quality slope | 0.02 per GW | Curtailment formula | Calibrated to 2024 actual ~2 TWh curtailment at 6 GW corridor. Formula: gridQuality = min(0.99, 0.85 + (corridorGW − 4) × 0.02). At 6 GW → 89% quality; at 15 GW → 99%. The fraction (1 − gridQuality) of post-BESS/pumped noon surplus becomes curtailment. No first-principles derivation — calibrated to match observed 2024–25 curtailment. |
| Corridor gas saving | 2.5% per GW | Panel 8 corridor €/tCO₂ | Modeller estimate from REE corridor expansion plans. Not from a dispatch model. |
| CO₂ factor (Spanish CCGT) | 0.40 tCO₂/MWh | All CO₂ calculations | 55% efficient gas turbine, EU emission factor. Standard. |
| Solar LCOE | €32/MWh | System cost | IRENA Spain 2025. Lowest in Europe. |
| Wind LCOE | €50/MWh | System cost | Spanish auction 2024-25. |
| Solar capex | €700M/GW | Panel 8 | BNEF 2025 utility-scale Spain. |
| Wind capex | €1,200M/GW | Panel 8 | BNEF 2025 onshore wind Spain. |
| Nuclear life-ext capex | €500M/GW | Panel 8 | OECD NEA 2024 LTO costs. 30-year extension assumed. |
| Field BESS capex | €150/kWh | BESS LCOS | BNEF Europe 2025. 4-hour systems. |
| Asset lives | 25 yr (RE), 30 yr (nuc), 15 yr (BESS) | Annuity calc | Standard. Nuclear extension 30 yr assumes Spanish 60-yr licence path. |
| Solar fleet CF (blended) | 12.7% | TWh per GW | 53.2 TWh / 48 GW / 8.76 = derived from REE 2025. |
| Wind fleet CF (blended) | 21.7% | TWh per GW | 62.8 TWh / 33 GW / 8.76 = derived from REE 2025. |
| Nuclear CF | 83% | Nuclear TWh | Spanish fleet average 2024. |
| Other firm sources | 7.6 TWh flat | Supply pool | CSP 2.4 + biomass 4.1 + coal 1.1 from REE 2025. Inflexible — does not respond to sliders. |
| CHP annual | 16 TWh, winter-shaped | Supply pool | REE 2025. Industrial heat-led shape (winter > summer). |
| Gas baseline (price-hours formula) | 45 TWh | "Hours gas sets price" card | Anchored to model default. Power-law decay with exponent 0.6. |
| Pumped charge efficiency | 80% RTE | Curtailment absorption | Standard pumped hydro round-trip. |
| BESS charge efficiency | 92% / 95% | Curtailment absorption | Standard Li-ion charge/discharge. |
Outside scope: The model is an annual energy balance with monthly disaggregation. It does not simulate hourly dispatch, merit-order clearing, firm capacity adequacy, transmission flow physics, wholesale price formation, or sector-coupling pathways. Those are the territory of full-resolution models (PLEXOS, Aurora, TIMES). For the question this tool addresses — which policy levers move which outcomes, and in what order of magnitude — annual resolution with calibrated dynamic relations is sufficient.
Grid stability / inertia note: The April 2025 blackout exposed a synchronous-mass shortfall under high-RE conditions. This is a separate concern from the energy balance modelled here. Spain is addressing it independently with synchronous condensers and rotating-mass devices ("flywheel"-class units), so the model does not encode an inertia floor on gas. A planner using this tool to set a low gas target should verify that the inertia/grid-forming buildout is on track in parallel.
Move sliders back to defaults to see this comparison. Numbers in italics are derived from model; non-italic are REE/Ember actuals.
| Metric | Model output (defaults) | REE / Ember 2025 actual | Δ | Comment |
|---|---|---|---|---|
| Gas peaker generation | ~45 TWh | 45.7 TWh metered (38.7 normalised) | −2% | Within rounding band ✓ |
| Solar generation | 53.2 TWh | 50.2 metered + 3 curtailed = 53.2 | 0% | Calibrated ✓ |
| Wind generation | 62.8 TWh | 58.8 metered + 4 curtailed = 62.8 | 0% | Calibrated ✓ |
| Nuclear generation | 50.9 TWh | 51.7 TWh | −1.5% | Within rounding ✓ |
| Curtailment | ~3 TWh | ~3 TWh structural (Aurora 2026 fwd) | 0% | Tuned to match ✓ |
| Hours gas sets price | 15% | ~15% (Ember 2026) | 0% | Anchored — formula not validated forward |
| CO₂ emissions (power) | ~18 Mt | 29.5 Mt (REE incl. all fossil) | −39% | Model only counts gas peaker; REE includes coal+CHP+CSP+islands |
| Net exports | 13 TWh (slider) | 13 TWh | 0% | Set by user. Modelled flat across months — real exports are summer-skewed; winter gas slightly overstated as a result. |
| Solar capture rate | 52% | 52% (Aurora 2025) | 0% | Anchored — base case validates only |
What is and isn't validated: The annual energy balance reproduces REE 2025 within ~2%. Capture rates and displacement rates are anchored to Aurora/Ember 2025 published values; their dynamic shape with demand follows an S-curve calibrated to grid physics. The €/tCO₂ ranking in Panel 8 (nuclear < grid/wind < solar at current demand) survives ±50% perturbation of every heuristic coefficient — it is a structural result of when each source generates relative to when gas runs, not an artefact of the constants chosen. The S-curve also captures solar's improving case as Spain electrifies — at 360 TWh demand with E/W orientation and storage, solar approaches wind on €/tCO₂. Direction is robust; magnitudes are indicative.