Solar power forecasting
A solar power forecast predicts the output of a photovoltaic plant hour by hour. Frontwatt builds it from irradiance scenarios and the physics of your plant, including panel orientation, temperature and inverter limits, and reports a range rather than a single curve.
Updated
What makes solar output uncertain
- Clouds. Output under a clear sky is easy to predict. The difficulty is cloud: when it forms, how thick it is and when it clears. Days with broken cloud have the widest ranges.
- Direct and diffuse light. Tilted panels receive direct sunlight, diffuse sky light and light reflected from the ground in different amounts. A forecast of total irradiance on a flat surface is not enough.
- Heat. Crystalline panels lose roughly 0.3–0.4% of their efficiency for every degree the cells warm up, so a hot, clear afternoon produces less than the irradiance alone suggests.
- Inverter limits. Many plants have more panel (DC) capacity than inverter (AC) capacity. On the brightest hours output is capped at the inverter limit, which is called clipping.
How Frontwatt forecasts a PV plant
- Irradiance scenarios. Each scenario from the global ensemble models, including the ECMWF IFS and AIFS ensembles, carries irradiance, temperature and wind for your site. Models that do not forecast irradiance are left out for solar plants.
- Sun position. We compute where the sun is for the middle of every hour, so the geometry is right at sunrise and sunset.
- Light on the panel. Irradiance is split into direct and diffuse parts and transposed to the tilt and orientation of your panels.
- Cell temperature and DC power. A standard photovoltaic model turns light and temperature into DC power, minus system losses.
- AC limit. Output is capped at inverter capacity. Night hours are zero by construction, and this is checked in every release.
- Calibration and limits. Metered history, when available, corrects systematic errors. Planned outages and curtailment are applied last.
In short: the physical chain is right from the first day if the plant data is right. The most common source of systematic error is wrong panel tilt or orientation, and metered output reveals it quickly.
What you need to provide
| Input | Why it matters | Needed |
|---|---|---|
| Location, DC and AC capacity | Geometry and the clipping limit | Yes |
| Panel tilt and orientation | Light on the panel surface | Recommended |
| Hourly metered output | Calibration and scoring | Recommended |
| Planned outages and curtailment | Keeps the forecast within real limits | When they occur |
Unlicensed plants in Türkiye
Hourly output of unlicensed solar plants is not published plant by plant in Türkiye. Frontwatt still forecasts them from plant data, and calibration starts once you share metered output.
What you get
- Hourly P10 to P90 ranges and the median for the rest of today, tomorrow and the day after.
- A new forecast each time a new weather model run is published.
- The day-ahead schedule with the lowest expected imbalance cost.
- Daily accuracy scorecards against your metered output.