Courtesy of the California Policy Center, commentary on the watergy nexus of covering aqueducts and canals:
Earlier this week, a respected YouTube channel covering the future of technology and energy released a video exploring the viability of solar panels covering California’s canals and aqueducts. The videos are hosted by Ricky Roy, a California-based mechanical engineer whose channel is called “Two Bit Da Vinci.” Reviewing his findings is an opportunity to offer some updates and perspective on the potential of photovoltaic energy in California.
Roy claims solar and battery storage capacity has exploded in recent years. He’s right. It’s grown so fast that even its proponents have seen their expectations surpassed. Worldwide photovoltaic capacity has increased by more than ten times in the last ten years, and the last decade has seen worldwide battery storage capacity increase by more than 150 times.
One of the leaders in this gallop into the electric age is California. The California ISO (Independent System Operator) website is a good place to get a visual impression of just how much growth we’re seeing in solar and battery capacity. On August 9, 2026, utility solar output exceeded 20 gigawatts from 9:30 a.m. till 3:30 p.m., six full hours, and later that same day, batteries discharged over 10 gigawatts into the grid from 6:45 p.m. until 8:30 p.m. It’s summer. Lot’s of sun. Of course! But consider the trend.
One year ago, on August 9, 2025, solar exceeded 20 gigawatts from 10:30 till 2 p.m., only 3.5 hours, and battery discharge only exceeded 9 gigawatts for about 30 minutes around 7:00 p.m.
Two years ago, on August 9, 2024, solar exceeded 18 gigawatts from 11:45 to around 2:15 p.m., only 2.5 hours, and battery discharge only exceeded 6 gigawatts for less than 30 minutes around 7 p.m.
Let’s go back one more year, to August 9, 2023. On that day, solar only broke 13 gigawatts from 10 a.m. to 3 p.m., and battery discharge only exceeded two gigawatts from 6:45 to 8:15.
There are better ways to explore this data, and when the California Energy Commission releases its Total System Electric Generation report for 2025, we’ll have a good look. But meanwhile, these anecdotes from the ISO website send a clear message: the days of lucrative price arbitrage by buying and storing the mid-day surplus and selling it back to the grid during peak demand around sunset are numbered. Battery discharge capacity in California has quintupled in just three years.
Peak grid demand in California is usually slightly over 40 gigawatts. Batteries can already cover 25 percent of that. If these installation trends continue, battery-supplied electricity in the evening will become a commodity, driving prices down to only around 25 percent more than mid-day solar electricity. That premium will cover — these are rough estimates — a charge/discharge loss of 10 percent, plus 10-15 percent to amortize declining battery productivity over its full service life, and 2 percent for transaction fees. Tack onto that as well about $0.03 per kilowatt-hour to pay down the loan to purchase and connect the batteries.
It would be presumptuous to assume California’s electricity prices will stabilize and drop even as we relentlessly electrify our economy. Here come more EVs and heat pumps, then electric trucks, more data centers, and, perhaps sooner than we expect, robots. There’s seemingly no end to it, and the state’s official goal of 500,000 GWH/year (up from 278,000 GWH in 2024) may be underestimating the need. But the price of solar/battery electricity is plummeting as its capacity soars. And then there’s enhanced geothermal, advanced nuclear, and natural gas generation with or without CO2 harvesting or sequestration (even with sequestration, it’s economically competitive).
What does this have to do with solar panels above California’s aqueducts and canals? Only this: Solar electricity is cheap, and the arbitrage margins realized by investing in storage are going to get thin. Successful solar will be achieved by keeping costs as low as possible. According to Roy, the installation cost per watt on land today ranges between $0.95 and $1.23. That sounds about right. He then reports that because of the structural requirements, installation cost per watt above canals is projected to cost between $1.90 and $2.46. This is a challenging disparity.
Roy also claims if all of California’s 4,000 miles of canals and aqueducts were covered with panels, it would generate 13 gigawatts and save 63 billion gallons of water by reducing evaporative losses from the canals by 85 percent. That’s 193,000 acre feet, which is a lot. But 13 gigawatts installed at $2 per watt is going to cost $26 billion, and if that same amount of photovoltaic capacity was installed on land, it would cost $13 billion. The savings by sticking to land would pay for the Sites Reservoir with enough left over to repair every aqueduct damaged by subsidence.
A lot of the economic analysis that investors have to evaluate involves subsidies and tax incentives. But ultimately, the economic cost doesn’t change. It’s either fully supported by the investors or it is socialized. One way or another, we all pay.