# Winter solar: capacity, costs and summer value

WattInfo, 22 September 2026. These are component and seasonal-energy comparisons. No matched hourly whole-system price has been established.

## Seasonal capacity

Assume 55 TWh winter demand including grid losses and 20 TWh natural-water hydro. The remaining 35 TWh comes from PV. Divide by the published modelled October–March yields: 296.9 or 257.8 kWh/kWp. The calculated 117.9–135.8 GWp motivates rounded 120/140 GWp cost examples. Additional storage losses and curtailment are zero in this energy screen; nuclear, wind, biomass and imports are omitted. Equivalent sites and continuous supply are unproven. The yield study uses 2004–2018 weather.

Installation CHF billion = (target GWp − 9.4986 GWp already installed at end-2025) × CHF/kWp / 1000. The 800, 1428 and 1911 unit prices are surveyed 2025 ex-VAT roof medians for different size bands, extrapolated as explicit all-at-one-price assumptions. They exclude finance, replacements, roof work, batteries and off-site networks. They do not form a feasible national site mix. The ageing sensitivity targets first-year-equivalent output with uniform builds, 30-year lives and 0.5% degradation; its target differs from nameplate capacity.

## Seasonal plant economics

Winter recovery price = (discounted plant expenditure − effective summer value × discounted potential summer MWh) / discounted potential winter MWh. The chart uses 60 years from first operation and a 5% real required return. The summer-value assumptions, zero and CHF50 per potential MWh, apply to all plants. Value is net of curtailment, unsold output, losses and transaction costs. Winter output is assumed fully useful. No matched hourly firming or off-site grid is priced.

Solar/wind replacement after 30 years is included. Solar has 0.5% annual degradation, upkeep including inverter replacement and a separate assumed 5% removal allowance. Nuclear reconstructs Axpo reference costs with construction carry, 85% output, 55% winter share, refurbishment, fuel and waste/decommission provisions plus a provisional insurance allowance. It is separate from the 2050 programme owner budget. The detailed assumptions and source catalogue retain differences in source dates and evidence strength. The 1.002 consumer-price conversion for Axpo 2024 figures is not a construction escalation forecast. Other mixed-vintage reference costs and project budgets are explicitly carried as constant-2025-price assumptions.

## Storage and networks

Battery size 40 GWh / 10 GW is an illustration, not a required system quantity. Costs assume CHF250–500 per usable AC kWh, 15 years, 5% return, 4% annual O&M including augmentation, 300 cycles, 85% efficiency and CHF30/MWh charging. Replacement is already recovered through the annuity; do not add it again.

Grimselsee receipt = capital recovery of CHF307m over 60 years at 5%, divided by one 240 GWh shift each year, plus CHF10/MWh upkeep and CHF30/MWh foregone summer sales. This prices a finite storage service, not new annual natural-water energy.

National network renewal and electrification investment must be separated from portfolio-specific reinforcement. Unknown whole-system costs remain null. The comparison contract and remaining gaps describe the needed hourly, geographical and commercial work.

## Reproduction

The downloadable JSON tables preserve the parent calculation outputs. The reproduction ZIP contains the calculation and independent validation scripts, their curated input/output files and this method. Original third-party documents and large hourly files are excluded; their dates, archive hashes and URLs remain in the source catalogues. Source-hash checks require those local archives. No new hourly dispatch was run for this chapter.