According to Quartz Japan has just built a 2.3 MW floating solar farm.
Why would they do this? Is it not cheaper to just stick the panels on land?
Well, Japan has a problem. There isn’t much land for doing anything. Here is how Clive James described the landscape of Japan:
From the air Siberia looks like cold nothing. The Sea of Japan looks like wet nothing. But Japan itself, at your first glimpse of it, looks like something. Even geographically it’s a busy place.
Immediately you are impressed by the wealth of detail – an impression that will never leave you for as long as you are there. Only a tenth of the land is useful for anything. The remaining nine-tenths, when you look down on it, is a kind of corduroy velvet: country so precipitously convoluted that the rivers flowing through it look like the silver trails of inebriated slugs. The useful tenth is inhabited, cultivated and industrialised with an intensity that boggles the Occidental mind. I had never seen anything like it in my life.
Depending on how you add it up, roughly 70 or 80% of Japan is forest and mountains. Only 12.5% of Japan is agricultural. A megacity like Tokyo takes up more than a little space. Put simply, there is very little room available for solar panels.
Let’s put some harder numbers on it.
The best way to understand the land requirements of solar power is to think in terms of power density, that is average output per unit land. I will use watts per square metre (W/m2) as the measure.
Japan Times tells us that this 2.3 MW solar farm covers 333 by 77 metres, and will produce 2680 MWh per year. This gives a power density of production of 11.93 W/m2. David MacKay calculated the power density of Kagoshima, a 70 MW solar farm, to be 8.6 W/m2.
We can estimate power density of Japan’s energy consumption. Using the most recent data from BP, I estimate power density of final energy consumption in Japan to be around 1.3 W/m2. This covers everything from electricity generation to heating a home. Some people like to focus on electricity generation alone, which is misguided given that fundamental goal is to decarbonize the energy system, not just one part of it.
So you would need to cover roughly 11% of Japan in solar panels to provide all of its energy needs.
[Yes, this back of the envelope calculation ignores the fact that the sun goes down and that 100% solar would likely require significant overbuild of capacity. It also ignores possible future improvements in efficiency.]
According to the FAO, around 80% of Japan is forest or agricultural. The Greater Tokyo Area takes up roughly 3.5% of Japan, and is not exactly an idea location for solar farms.
This takes up back to Clive James’s observation. 90% of Japan is probably not available for solar panels. Hence the construction of solar farms on water. Anything close to 100% solar, then, will require Japan to cover almost all available areas in solar panels. This is not easily done.