Ian Welsh tweeted, “Japan HAS to import energy.”
Is he right? It turns out that photovoltaic changes things; Japan can sustain its current energy consumption on about 5% of its land area, using plain cheap photovoltaic panels. It will probably get the majority of its energy from photovoltaic panels starting in the mid-2020s, based on current growth rates.
Wikipedia says Japan used 477.6 Mtoe in 2011, which is 19.996 exajoules, a number suspiciously close to 20, or 630 GW average. The accompanying chart shows that the number is closer to 21 EJ, or 670 GW, roughly level since 1995. Solargis’s solar resource chart gives numbers of global horizontal irradiance (GHI) from 1200 to 1600 kWh/m²/year; a rough average might be 1400, although Wikipedia gives somewhat higher numbers of 4.3 to 4.8 kWh/(m·day), which works out to 1600–1800 kWh/m²/year. Solargis also offer a “potential PV electricity generation” map for optimally-tilted free-standing crystalline silicon modules, ranging from some 900 to some 1600 kWh/kWp, with most of the land around 1200.
Japan’s area is 378000 km², which gives us a total GHI of about 500 PWh/year, or 60 TW. With 16%-efficient (polycrystalline Si) horizontal solar cells, this gives us 10 TW, which is about 15 times Japan’s total national marketed energy consumption.
That is, Japan could satisfy its current energy demands with about a fifteenth of its total solar resource, occupying perhaps 5% or 10% of its land area. To supply these 670 GW with an unremarkable solar photovoltaic capacity factor of 20% would require 3.3 TW of solar panel nameplate capacity (TWp); currently installed at the end of 2016 is 42.75 GW, about 78 times smaller. Installed PV capacity rose from 3.618 MW at the end of 2010 to 42.75 GW at the end of 2016, a factor of 11.82, an exponential growth rate of 51% per year. At this exponential rate, Japan would need another 10.6 years (mid-2027) to reach 100% of energy from domestic solar photovoltaic. Presumably the growth rate will slow earlier than that, perhaps around 2023.
At current photovoltaic prices of €0.53/Wp (May 2017, Japan/Korea market), 3.3 TWp is 1.7 trillion euros. Japan’s nominal GDP is currently US$4.73 trillion per year, although only has US$697 billion in exports (€608 billion at US$1.146/€). So this represents an investment of about 2.8 years of Japan’s exports or about 0.41 years (5 months) of Japan’s GDP. Currently Japan pays a very high tariff of ¥42/kWh for solar, which, at ¥114/US$, is US$0.37/kWh or €0.32/kWh. So the full €1.7T investment will be paid for by the next 5.3 trillion kWh (19 EJ) of solar energy consumed in Japan.
That’s only the cost of the panels, but at this point the levelized cost of new PV plants is under US$1 per peak watt, even including the uncertainty premium to investors. That is, the cost of the whole plant is less than twice that of the panels. We can expect that cost to continue coming down.
There are some other obstacles to going majority-solar, such as utility-scale energy storage for baseload power (photovoltaic panels produce no significant power at night) and producing liquid fuels for transport. I expect them to be solved once the incentive of abundant but intermittent and stationary solar energy is there, which will probably require reducing the high solar tariff.
So, although Japan can definitely afford to switch to mostly photovoltaic, it can’t do it much sooner than 2027, because it doesn’t have enough exports.
utility-scale