California’s 1.6-megawatt solar-canal pilot is operating, but the 63-billion-gallon water figure remains a statewide model estimate that the two sites are only beginning to test.
By The Long View · Edited by Lachlan Brown
Published September 14, 2026 · How we edit
Follow Science Blog on Google
California’s Project Nexus finished two solar-canopy installations over working irrigation canals in 2025. Together they provide more than 1.6 megawatts of generating capacity while testing whether one piece of infrastructure can produce electricity, conserve water and avoid taking farmland out of use.
The much larger number attached to the project, 63 billion gallons of water a year, is not a measured saving from those two sites. It came from a 2021 computer model of covering roughly 6,350 kilometres of California’s major canals, rounded to 6,400 kilometres in the title.
This is one study and one demonstration project, not settled evidence that every canal should carry solar panels.
Project Nexus sits in the Turlock Irrigation District in California’s Central Valley. According to the district’s project page, the system covers sections at two locations and supplies more than 1.6 megawatts to TID’s power portfolio.
The first installation crosses a canal about 20 feet wide. Its panels were fully mounted in March 2025. The second crosses a much broader channel, about 115 feet wide, and was completed and energized at the end of August 2025.
That difference is part of the experiment.
A structure over a narrow waterway does not answer every question about longer spans, wind loading, support placement or access for canal maintenance. Building both gives the partners operating examples rather than relying only on engineering drawings.
The California Department of Water Resources awarded $20 million to the pilot in 2022. TID, UC Merced and developer Solar AquaGrid are the other principal partners.
Brandi McKuin and colleagues at UC Merced and UC Santa Cruz published the underlying study in Nature Sustainability. It used regional hydrologic and economic simulations to compare solar panels over canals with conventional arrays built on nearby land.
The model covered about 6,350 kilometres of major California canals across different climates, sunlight levels and water values. It estimated average annual evaporation savings of 39,000 cubic metres per kilometre, with uncertainty of plus or minus 12,000 cubic metres.
Scaled across the modeled network, the result was about 63 billion US gallons, equivalent to roughly 240 billion litres, conserved in an average year. A UC Merced explanation of the paper compared that volume with the residential water needs of more than two million people or the irrigation needs of 50,000 acres of farmland.
Those are scale comparisons, not delivery plans. Water retained in a canal system is not automatically allocated to two million residents, and actual savings would depend on which canals were covered, local weather, water depth, flow and the fraction of each channel shaded.
The study also projected about 13 gigawatts of solar capacity if the suitable network were built out. Project Nexus is 1.6 megawatts, about eight thousand times smaller, so it should be understood as a test bed rather than a miniature proof of the statewide total.
Photovoltaic cells generally become less efficient as their operating temperature rises. Air immediately above open water can be cooler than air over sun-baked ground because evaporation absorbs heat. Panels spanning a canal may therefore convert sunlight slightly more efficiently than comparable equipment over dry land.
The exchange is reciprocal. The panels shade the canal, reducing solar energy available to drive evaporation, while evaporation that still occurs helps cool the panels. Shade may also slow the growth of aquatic weeds that obstruct water flow and require mechanical or chemical control.
The cooler-microclimate claim is physically plausible and supported by earlier canal-top installations, but the gain is not one fixed percentage for every site. Panel design, clearance above the water, wind, humidity and temperature all affect performance.
California did not invent the configuration. ScienceBlog has reported on Gujarat’s early canal-top solar project, completed in India in 2012. That installation helped establish that electricity could be generated above an operating irrigation channel while leaving neighboring land available for agriculture.
A regional simulation cannot capture every operating constraint. Canals vary in width, orientation, bank material, access roads, sediment load and proximity to power connections. Some need frequent maintenance or cross equipment routes that an overhead frame could obstruct.
Project Nexus therefore includes sensors and monitoring for electricity output, evaporation and water conditions. The University of California reported the system online, generating power and data after the second site was completed.
That distinction between model and measurement is central. The 2021 study motivated the pilot; it did not measure what Project Nexus would save before the project existed. Multiple seasons of site data are needed to test how much water the canopies retain and how their panels perform through heat, wind, irrigation changes and maintenance work.
Two sites also cannot represent 6,400 kilometres of infrastructure. Their results can validate mechanisms and expose practical problems, but statewide estimates will still require information from a wider range of canals.
Putting panels over water saves land, yet it requires longer and stronger supports than an ordinary ground-mounted array. Construction crews must work around a functioning canal, and operators must retain access for inspections, weed removal, sediment management and repairs.
The Nature Sustainability model found that the financial value of conserved water, avoided land, potentially lower weed-maintenance costs and electricity could outweigh the added price of the supporting structure. Its estimated net present value for canal-top solar exceeded nearby ground-mounted solar by 20 to 50 percent under the scenarios tested.
That is not a universal price guarantee. Financing rates, steel costs, land prices, water value and grid access vary, while not all benefits accrue to the same organization. A water district may pay for structures while a power buyer receives much of the electricity value unless contracts distribute those gains.
Project Nexus turns an appealing systems model into equipment that engineers and canal operators have to live with. Its next contribution will be measured performance: how much water the two sites actually save, how much extra electricity cooling delivers and what it costs to keep the canals working beneath them.
Follow Science Blog on Google
Add us as a preferred source to see more Science Blog reporting in Google.
Standards
Articles on Science Blog are edited and fact-checked before publication. We use AI tools in the newsroom. See our editorial standards and masthead.
A Science Blog column
The Long View steps back. A single finding rarely means much on its own, so this desk connects new research to what came before it, across physics, biology, medicine, and everything in between.
Clear science for curious people
A Brown Brothers Media publication. All rights reserved © 2026.
Clear science for curious people — free PDF, then our science email.
Also available via RSS.

source