Is desalinating enough water to irrigate an entire valley less viable than nuclear greenhouses? I'm not so sure. Both would be extraordinarily expensive
desalination cost varies, but for the London desalination plant it's £1.6 per litre[1]. I am sure that's on the higher end, but let's just work through the math.
1 mm of rain on 1 ha of land is 10 000 litres or 10 cubic meters of water. So to replace 1mm of rain in Britain, costs £16. Average annual rainfall in UK ~1150 mm. So if you had to replace the entirety of it from desalination, you would spend like £18,400.
If you are growing wheat, you get like 8 or 9 tons per hectare and you sell if for like £230 per tonne (these are very rough numbers) so your total revenue is like £2,000 per hectare. Out of that, your profit margin is probably under 10%
So you will be loosing money you have to replace more than a couple of % of annual rainfall from desalination, and if you grew wheat with entirely desalinated water it would cost 10X as much as 'normal' wheat does.
Now, typically you only irrigate high value crops, and wheat is a low value crop. Also you probably don't need the entire annual rainfall, and there might be cheaper desalination elsewhere. You can re-do the math with other crops and other desalination costs.
But fundamentally price of desalinated water is too high for use in agriculture.
Not at market energy prices, that's for sure. One possible exception would be if you coupled the desalination system to a solar farm and you only ran it at midday when generation exceeded demand. If you're getting the electricity "for free", it would change the calculation substantially.
