The installed base is now measured in gigawatts, not demonstration plots

Global agrivoltaic capacity reached an estimated 14 GW by the end of 2023, according to review data compiled from national inventories, with China's photovoltaic greenhouses and Japan's long-established "solar sharing" arrays accounting for the bulk of that figure. Japan has been deploying dual-use arrays since 2004 and now counts several thousand registered agrivoltaic plants, while the U.S. Department of Energy's InSPIRE programme has catalogued more than 300 agrivoltaic sites across the United States, a number that has roughly tripled since 2018.
The pipeline matters more than the installed base. NREL modelling published in 2023 found that co-locating PV with agriculture across existing U.S. farmland could technically supply a share of national electricity demand equivalent to more than 20% of annual generation while occupying less than 1% of agricultural land. Even if only a fraction of that is built, it reframes land use: solar developers facing county-level moratoria and agricultural opposition now have a defensible answer that keeps the land in production rather than retiring it for 30 years.
Commercial forecasts track the same curve. Several market research houses put the agrivoltaics technology and services segment in the range of $6–9 billion by 2031, growing at a compound annual rate above 15% from a 2024 base. Those numbers should be treated cautiously because definitions vary — a sheep-grazed conventional array and a 5-metre stilted wheat array are counted in the same category by some analysts and not by others — but the direction is unambiguous. In 2024, for the first time, more than one European country ran a national tendering round reserved exclusively for agrivoltaic projects.

What dual land use actually delivers in the field

The headline benefit of agrivoltaics is the land equivalent ratio (LER), which compares the yield of a dual-use system against the same land used separately for energy and agriculture. Well-designed projects report LERs between 1.2 and 1.8, meaning the combined output is 20–80% higher than mono-use. The 2016 Heggelbach pilot in Germany — a 194 kWp stilted array over arable land run under the APV-RESOLA research programme — produced LERs up to 1.86 in favourable crop scenarios, though potato yields under the array fell by around 18% that season.
Water is often the bigger prize. A widely cited study from the University of Arizona published in Nature Sustainability found that cherry tomatoes grown under PV in the Sonoran Desert produced roughly double the fruit of open-field controls while the panels' shading cut irrigation demand on the cultivated area by around 45–50%. The mechanism is straightforward: modules intercept part of the incoming radiation, lower soil and canopy temperature, and reduce evapotranspiration. In water-stressed regions, that water saving is worth more than the small yield penalty in shade-intolerant crops.
Microclimate effects are crop-specific and cannot be assumed. Lettuce, spinach, chard, peppers, berries, hops, saffron, agave, mushrooms and forage grasses are consistently reported as neutral to positive under partial shading. Maize, wheat and other C4 or high-light cereals generally lose yield under dense arrays unless row spacing, module elevation and orientation are tuned to keep daily light integral above roughly 70–80% of open-field levels. Growers should assume every crop needs a season of local measurement before scaling.

The economics hinge on revenue stacking, not on module cost

Agrivoltaic CAPEX runs 15–60% above a conventional ground-mount of the same nameplate capacity, depending on configuration. Elevated structures with 4–5 metre ground clearance require longer piles, heavier steel, deeper foundations and higher wind loading calculations — in Europe, a stilted array typically lands at €1,000–1,400/kWp against €600–800/kWp for a standard fixed-tilt plant. That premium has to be paid back by agricultural revenue, policy support, or a land cost that is effectively zero because the farmer remains a partner rather than a landlord.
On the revenue side, the array keeps generating. Bifacial modules mounted at height benefit from rear-side irradiance reflected off crops, bare soil and grass, with rear gains in the 5–15% range depending on albedo and elevation. Vertical bifacial rows oriented north–south produce a characteristic twin-peak output profile with generation concentrated in morning and evening hours, which can be more valuable on markets with high midday solar penetration. In France and Germany, agrivoltaic tenders have cleared at tariff levels above comparable open-field auctions, reflecting the recognition of higher build cost.
Livestock integration is the cheapest entry point. Sheep grazing under conventional 1.2–2 metre ground-mount arrays requires almost no structural premium and replaces mechanical mowing, which typically costs a utility-scale asset €200–400 per hectare per year. Solar grazing has spread to more than 5 GW of U.S. capacity in some form, concentrated in the Midwest and Mid-Atlantic, with shepherds contracted by asset owners under multi-year vegetation management agreements. For asset owners comparing revenue models, dual use is less a premium product than a hedge against land access, community opposition and escalating O&M costs.

Four configurations, four different engineering problems

| Configuration | Ground clearance | Typical row pitch | Best-fit production | CAPEX premium vs ground-mount | Key constraint |
|---|---|---|---|---|---|
| Stilted elevated array | 3.5–5.0 m | 6–12 m | Wheat, potatoes, forage, vineyards | +30–60% | Structural wind load, tractor clearance |
| Vertical bifacial rows | 2.0–2.5 m height | 10–15 m | Grassland, cereals, pasture | +15–25% | Low specific yield, high land take |
| Solar greenhouse roof | 3–6 m | n/a | Tomatoes, lettuce, herbs | +40–80% | Humidity, glass cost, crop lighting |
| Ground-mount with grazing | 1.2–2.0 m | 3–5 m | Sheep, poultry, pollinator mix | +0–5% | Grazing pressure, water access |
The table above is not a hierarchy. A developer choosing between a vertical bifacial layout and a stilted tracker array is choosing between two different businesses: the vertical layout maximises cropped area and keeps machinery access simple, while the elevated array maximises energy density and suits premium horticulture. Australian and German trials with vertical bifacial systems have recorded higher land equivalent ratios than tilted equivalents in pasture, but 20–30% lower kWh per installed watt.

Policy is now the decisive variable

France's 2024 agrivoltaics decree is the most prescriptive regime in Europe. It requires that agricultural production remain the primary activity on a dual-use plot, caps panel coverage of a given parcel, and obliges developers to demonstrate at least one measurable agricultural service — yield improvement, climate adaptation such as frost or drought protection, hazard mitigation, or improved animal welfare. Projects that fail the test revert to standard ground-mount rules, which under the 2023 APER law effectively removes them from agricultural land eligibility.
Germany took a market route instead. The Renewable Energy Sources Act introduced dedicated agrivoltaic tenders from 2023, with higher ceiling prices than open-field rounds and an eligibility definition that requires land to remain agriculturally productive, with a maximum loss of usable area for arable configurations. Fraunhofer ISE has estimated Germany's technical agrivoltaic potential in the high hundreds of gigawatts, though the realistic near-term number is one to two orders of magnitude smaller and constrained by grid connection rather than land.
In the United States, policy is fragmented across states. Massachusetts, New Jersey, New York and Illinois have either incentive adders or siting standards that recognise dual use, while several Midwestern counties have moved in the opposite direction with restrictive ordinances. Japan's framework, meanwhile, is built on the original solar sharing principle of continued agricultural output, with crop yield reporting obligations attached to feed-in tariff approvals. For any developer underwriting a project, the practical implication is that agrivoltaics is a compliance product in some jurisdictions and an optional optimisation in others.

What to specify before you sign the EPC contract

Agricultural machinery access drives the structural design more than module technology does. A standard tractor with a sprayer needs at least 4 metres of vertical clearance and 6–9 metres of inter-row gap to work efficiently; the difference between a 4-metre and a 5-metre clearance on a 20 MW project can swing steel tonnage by double-digit percentages. Cable routing, tracker drive shafts and combiner boxes all need to be positioned so that ploughing, harvesting and grazing can continue without service interruptions, which usually means burying all DC cabling below ploughing depth.
Monitoring requirements are heavier than on a conventional plant. Agrivoltaic assets typically need soil moisture and temperature sensors, reference crop plots in both shaded and open zones, and irradiance measurement at array height to separate energy underperformance from agricultural underperformance. The industry standard for PV monitoring, IEC 61724-1, provides the framework, but most agrivoltaic contracts add bespoke yield-tracking clauses that tie lease or revenue-share payments to verified crop outcomes — and those clauses are meaningless without a well-instrumented control plot.
Storage and grid interaction deserve early attention, because dual-use sites are often built where the grid is weakest. Co-locating a battery storage system behind the meter lets the array shift its vertical or elevated-array output into evening price peaks, which is exactly where agrivoltaic generation profiles are strongest. Developers should model the co-located case from the start rather than retrofitting a DC-coupled battery three years into operation, since the inverter architecture and site layout differ substantially.

Commercial real-world examples worth studying

The Lake Constance region of southern Germany remains the reference site for stilted arable agrivoltaics, with the original Heggelbach array followed by larger commercial builds on the same structural logic: steel posts on 8–12 metre centres, bifacial glass-glass modules, and full agricultural machinery access. Output from these sites typically reaches 85–90% of a comparable open-field plant per installed watt, while the land continues producing silage, grains or vegetables under contract with regional buyers.
In the U.S. Midwest, the dominant model is grazing rather than cropping. Utility-scale assets of 100–300 MW contract with sheep farmers who manage vegetation across thousands of hectares, reducing mowing costs and improving soil carbon and pollinator habitat. Illinois and Minnesota have coupled these arrangements with pollinator-friendly seed mixes; Minnesota's standard, developed with university and utility input, has been applied across hundreds of megawatts and is now referenced by developers in neighbouring states as a de-risking tool for permitting.
Southern Europe is pursuing the same logic for permanent crops. Olive groves and vineyards with elevated or semi-transparent arrays have shown lower evaporation and heat stress during summer peaks, protecting yields in years that would otherwise be loss-making. For investors, these projects are attractive because they stack agricultural revenue — which is inflation-linked and uncorrelated with power prices — on top of a 25–30 year PPA. That correlation benefit is increasingly valued by infrastructure funds, and it is the reason agrivoltaics has moved from sustainability reporting to investment committee agenda.

Design considerations vary by land type, and mistakes are expensive

A common failure mode is treating agrivoltaics as a standard array with taller piles. Row spacing selected for maximum energy yield, rather than for the specific combine, harvester or tractor that will work the field, is the single most frequent cause of project abandonment. On arable land, design teams should begin with the machinery specification and work backwards to tracker pitch and torque tube height, accepting a 5–10% energy density penalty as the cost of a working farm.
The second failure mode is contractual. Revenue-sharing agreements between the asset owner and the farmer need to define who bears yield risk from shading, hail, soiling and panel cleaning runoff, and who pays for damage from machinery. Insurance policies written for conventional ground-mount plants rarely anticipate a tractor striking a module edge at 4 metres, or livestock rubbing against junction boxes. Both are manageable, but only if they are priced before financial close rather than negotiated after a claim. For owners running mixed portfolios, standardising the inverter and monitoring architecture across agrivoltaic and conventional sites simplifies O&M considerably.

Where the market goes from here

Three forces will shape the next five years. First, grid capacity rather than land availability is becoming the binding constraint in mature markets, and agrivoltaic projects often sit where interconnection queues are shorter — rural agricultural feeders with existing distribution infrastructure. Second, module efficiency gains reduce the land area needed per megawatt, slowly lowering the structural cost premium per unit of agricultural output. Third, the emergence of vertical bifacial and semi-transparent module formats gives designers a wider toolkit for matching light transmission to specific crops.
The market is also consolidating around a smaller number of proven configurations. Most new capacity falls into one of four buckets: elevated arable arrays, vertical bifacial pasture systems, solar greenhouses, and grazed conventional plants. Developers weighing a first agrivoltaic project should compare these against site-specific factors — crop type, machinery, water availability, local policy — rather than against a generic benchmark, because the spread between good and bad agrivoltaic design is far wider than in conventional solar. Additional project references and structural options are available via utility-scale project references and through DLXN's product portfolio for high-clearance and bifacial mounting applications.
What is no longer in question is whether dual land use works at scale. It does, in a specific set of conditions, with a specific set of design rules. The projects that fail will fail for the same reasons solar projects have always failed: underestimating the site, the soil, and the people who work it.