Remember the “Solar Freakin’ Roadways?” The idea was to pave roads with solar cells that would produce electricity for the grid. The power could also be used for LEDs that could serve as traffic lights and warning signals, and the power could melt snow and ice to boot. I wrote about this in 2014 and at the time I was not impressed with the arguments that were being made.

As many people pointed out – there are huge practical problems with this concept. The main problem is that you would have to engineer the solar cells to withstand traffic and all the abuse that roads take. This would likely degrade their efficiency over time, and also make upkeep very expensive. For the money you would be far better off simply installing solar panels on rooftops, or even suspending them over roads or wherever that might make sense.

I wasn’t entirely negative. I liked the idea in theory, just thought there were serious practical hurdles and we need some real-world testing. I also said this:

“It’s also possible that such panels might find a niche use, but not be cost effective for our entire highway infrastructure. For example, they may find a market for private driveways. I would like not to have to plow or shovel my driveway, and it may reduce my energy bills a bit. Also, a driveway takes a lot less abuse than a highway.

Or perhaps parking lots, bike paths, or playgrounds may find a use. Perhaps they will be better for small back roads than highways, or the other way around, or they will be perfect for cities or sidewalks. Maybe airport runways might justify the cost.”

Well now it’s 5 years later, and we have some real-world testing. The largest test was in France, which two years ago installed a 1 kilometer stretch of solar roadway. Le Monde reports that the experiment was a complete failure. The roadway is deteriorating rapidly. A large section had to be completely demolished. Panels come loose and then are broken or dislodged by traffic. Leaves and other debris cover the surface. Also:

On top of the damage and poor wear of the road, the Normandy solar track also failed to fulfill its energy-production goals. The original aim was to produce 790 kWh each day, a quantity that could illuminate a population of between 3,000 and 5,000 inhabitants. But the rate produced stands at only about 50% of the original predicted estimates.

That 50% has declined by another 50% in the second year. The efficiency is rapidly waning as the road ages. All this means that the solar road is not cost effective. It does not produce enough energy to make the cost of construction and maintenance worth it. Further, the road surface is noisier than traditional paved road, and this resulted in authorities lowering the speed limit on the road.

There was a second solar road installed in the US, in Idaho, which also completely failed:

Roughly 25 out of 30 panels installed in a prototype solar road in Idaho broke within a week, after the project received $3.9 million in funding and 6.5 years of development.

The prototype appears to be plagued by drainage issues, poor manufacturing controls and fundamental design flaws.

You might argue that these failures are due to execution, not concept, and that may be true. However, they did fail in precisely the way predicted by skeptics. At the very least it means any practical solar roadway would have to be engineered to be much more resilient without significantly increasing the cost. This seems unlikely anytime soon, and it seems even more unlikely that there will be much enthusiasm for investing in this technology given these early results.

Sometimes a technology idea is just flawed, and will never see the light of day. Other ideas, however, may have merit but are just impractical given current technology. I tend to think that solar roadways are in the former category rather than the latter, but this is hard to predict. It may take the development of a new material, one that is cheap and rugged, to make the concept viable. It’s like the space elevator, the idea of having a cable extend from geosynchronous orbit to the ground. If we could get this to work, it would be a cheap and efficient way to get stuff into space. However, there are massive practical hurdles, and we simply do not have the material necessary to construct such a thing. It’s possible the space elevator will never be practical, despite the fact that it’s a cool idea.

There is some good news for solar road enthusiasts, however – Solaroad has installed a solar bike path in Holland that is doing much better than the test roads.

They had originally hoped to produce somewhere between 50 and 70 kWh per square meter per year, the first year actually yielded 73 kWh per square meter per year, and the second, with improvements in design, 93 kWh per square meter per year. However, with wear the efficiency dropped to the predicted range. For reference the average US home uses around 10,000 kWh per year, so you would need about 130 square meters to power a typical home.  That is the size of a driveway that is about 5 meters wide and 26 meters long. There are other variables as well, such as shading and orientation to the sun.

There is some wear on the Solaroad bike path, and efficiency did decrease over time, but they say they are improving the technology each year, specifically the top protective layer. They are also now using flexible solar panels. These have lower initial efficiency, but are much more resilient to use than the rigid silicon panels.

This kind of application makes much more sense – use it on a surface used by bikes and people walking, rather than cars and trucks. Develop the technology with this easier task, which may bootstrap it to more rigorous applications. Even if not, there are lots of surfaces where the application may be cost effective. Pick this low hanging fruit first, before going right to the most difficult application of actual roads.

The reason I predicted this outcome in the first place is partly because it fits a general pattern. Often a new technology promises to revolutionize the way we do things, but in reality they find only a niche application. Microwaves did not change the way we cook as initially promised, but is a very useful heating tool. Segways did not change the way we move around cities as promised, but are valued by mall cops and the like. Yet other technologies, like cars, do fundamentally change the way we do things. The challenge is in predicting which path a new technology will take (or the third path to total oblivion). Most new technologies that find a use don’t completely change our infrastructure or our lives, so that is a safe bet. And in this particular case it just made sense that a non-road surface would be a much more practical application.

This also gets back to my earlier discussion from two months ago – there is no one energy solution. Stop trying to find one solution to all our energy infrastructure needs. This is a flawed approach. We want our energy infrastructure to be as low carbon as possible, but otherwise we want it to be as cost effective, safe, and convenient as possible. This means picking the low hanging fruit for each type of energy source for each location and application. If there is a specific location and application for which a solar ground surface is the most cost effective and convenient application, then use it. But don’t force this application when other better options are available, just because it’s novel.

The above three experiments reinforce my earlier prediction – solar roadways will not come into general use on actual roadways, but may find a niche use on other less demanding surfaces where their use makes sense.