Panels cut crop water loss by a fifth in French agrivoltaic trials
Serguey Asael Shinder2 min readFrench developer TSE has published agronomic monitoring from its agrivoltaic canopies, and the numbers are specific enough to argue with. At Chadeleuf in Puy-de-Dôme, winter forage barley grown under the panels reached 90% to 96% of the yield in a control plot without them, while seasonal evapotranspiration fell by 20%. Grain specific weight — a quality measure — came in 3% above the control.
The instrumentation is the reason the figures are usable: 18 radiation sensors, 14 weather stations and 14 soil temperature and moisture probes across the site, a harvest taken with an experimental combine by the research company Antédis, and monitoring conducted and validated by the Puy-de-Dôme Chamber of Agriculture and the Limagne centre for technical agricultural studies.
The protective effect concentrates where it matters. At the hottest part of the day, air temperature at crop level was 4.1 °C lower beneath the canopy, and evapotranspiration in those periods dropped by 2.15 mm per day. On the coldest nights it ran up to 3 °C warmer underneath. Measurements at TSE's Brouchy site in 2025 and 2026 showed the structure narrowing the average daily temperature range by about 1.5 °C, and the same effect appeared at Chadeleuf this year. At a separate site, maize under the canopy yielded 19.6% more in an exceptionally hot, dry season.
TSE notes that every scenario met the French decree of 8 April 2024, which requires yield loss under an agrivoltaic installation to stay below 10%. The report is at https://www.pv-magazine.com/2026/09/18/french-agrivoltaic-trials-cut-water-loss-lift-maize-yields-by-19-6/.
What it means
The interesting result is the variance, not the mean. Barley lost a little; maize in a hot dry year gained a fifth. A canopy that costs you 4-10% in a normal season and hands back 19.6% in a bad one is not primarily a yield device — it is insurance against the tail, bought with panels that earn money the rest of the time. That is a different financial product from "solar plus farming", and it is priced differently.
A 10% statutory threshold turns agronomy into compliance. The French decree sets the number the trial has to beat, so every French agrivoltaic project now needs exactly this kind of instrumented control plot. Expect the measurement apparatus — sensors, weather stations, a validating chamber of agriculture — to become a standard line item, and expect the published results to cluster just inside the threshold, because that is what thresholds do.
And the water number may outrun the electricity number. Twenty percent less evapotranspiration over a season, in a region that has had water restrictions, is an argument that does not depend on the power price at all. Where irrigation is capped rather than merely expensive, shading may end up being sold as a water technology that happens to generate.