[cc438]

Land utilization is the issue. The largest solar farms in the US requires an order of magnitude more land to produce the same amount of power as a relatively average nuclear plant.

The largest solar farm in the US, the Topaz Solar Farm in California, covers 25sq/km has a peak generation capacity of 550MW. The source uses the median of the 59 nuclear plants in the US to arrive at its 3.36sq/km (1.3sq/mile) per 1,000MW figure but the largest nuclear plant ever built, the Kashiwazaki-Kariwa plant in Japan, produces ~8MW on just 4.2sq/km of land which makes it twice as efficient in terms of land use as the median US plant.

That places solar at 13.75 sq/km per 1,000MW compared to nuclear's 0.525sq/km using the best case scenario for the density of existing sources. That doesn't even account for the variable output of the solar plant versus the consistency of nuclear generation. Solar is ultimately far too land hungry to ever serve as the world's primary source of electrity.

Source: http://www.nei.org/CorporateSite/media/filefolder/Policy/Pap...

[londons_explore]

More importantly, with fusion, energy could become significantly cheaper allowing us to do things which weren't economically viable.

Make everything out of environmentally friendly, lighter and long lasting aluminium rather than steel + concrete.

Have street lighting that is as bright as sunlight over entire cities, allowing us to no longer be dependant on time of day.

Growing food crops indoors with artificial light, saving massive amounts of land.

Desalination to make all our water. No longer extracting it from rivers and aquifers. A cleaner, less polluted water source.

A single energy source - today we use petroleum (cars), natural gas (cooking and heating), electricity (lighting etc.), diesel (transport) and many more in industry. All of those need distribution networks. A single cheap energy source could coalesce a lot of infrastructure.

[dpark]

> Have street lighting that is as bright as sunlight over entire cities, allowing us to no longer be dependant on time of day.

Please, no. Some of what you said I'm onboard with, but definitely not this. The human impact of perpetual day is large and negative. The negative impact of massive artificial lighting on astronomy is also quite massive.

[motoboi]

I made me curious about something.

Will the heat from all that electricity generation affect the environment in significant ways?

At which energy usage scale humanity will be restricted by the planet environment capacity to deal with the generated heat?

[londons_explore]

If you have enough cheap energy, you could refrigerate earth and send excess heat to space.

I think we're a long way from that yet though.

[philipkglass]

Until recently the limiting factor on utility scale solar was high costs. Now intermittency looks like the most important constraint. Land availability is not the bottleneck for the US or most other countries. There are some countries that have a dense enough population and/or low enough insolation that land availability is a significant constraint on solar deployment (Japan, Taiwan, UK, Belgium...) but that's not common.

The real annualized power of the Topaz Solar Farm comes to 6 megawatts per square kilometer, using its 2015 generation total of 1,301,337 MWh. At that areal productivity[1], if you could ignore intermittency, it would take ~95,000 square kilometers to supply the 2013 US electricity demand of 4,986,400,000 MWh.

The US Corn Growers Association estimated that 27% of US corn grown in 2011 was for ethanol, and corn covered 92 million acres:

http://www.ethanolproducer.com/articles/8611/world-of-corn-r...

Converting 27% of 92 million acres to square kilometers, the US is using about 100,500 km^2 to produce ethanol.

But you can't ignore intermittency, so (barring much improved storage, which may happen but can't be assumed) solar PV is not a drop-in replacement for nuclear power.

[1]Other areas of the US get less sun than central California, but newer farms with more efficient panels and single axis tracking increase areal productivity significantly over the fixed-tilt CdTe construction of Topaz, so I think that 6 MW/km^2 is a reasonable number to use.