"I see that as an upper limit that may well soon be reached. Remember that demand is about to double... unless there are a lot of cancelations."
Again, so? In the context of energy prices set by generation from fossil fuel sources in the £40-50/MWh range, the uranium price would have to increase by more than an order of magnitude.
One small note - there's a technology change from Generation II to Generation III plants that will largely offset something like a doubling. One of the clever tricks is the use of "burnable poisons" to manage fuel reactivity. It means you start with a higher enrichment ([perhaps 4% insterad of 3%) but with the fuel doped with neutron absorbers. They "burn" and the net result is, it's possible to maintain reactivity with fuel burn-ups in the 8% region, instead of <5%.
"Add to this that the cost of solar power is falling every year... and that the US calculations already suggest that, there at least, it's cheaper than nuclear generation
Well, no. Because you've, as I explained, misunderstood what that "overnight cost number is.
The error this time isn't on the financial side - it's on the fact (check back) it's based on "nameplate" capacity. I.e the MAXIMUM rated output.
You've then got to include the effect of things like capacity factors. for example, with a wind plant, alhthough the nameplate cost may be (using real numbers from the London Array) about £2.5Bn/GW, the capital cost still needs to be recovered from actual output sales. So, if that plant will then run at 30% capacity factor, the actual figure needs to be adjusted - to what I make as £8.7Bn/GW.
You have to size a solar array to make "x" kw at noon, on a clear day. But it's not going to do that - allowing for darkness, oblique solar incidence, clouds, etc. you'd be hard-pushed to see a capacity factor of over 2--25% - so as judged against saleable output, you need to increase that cost by a factor of 4 or 5 to get a representative number.
And, if anything, I'd argue for further loading - because the generation isn't "dispatchable", i.e available when needed as opposed to availability being determined by time of the day, weather, etc.
".. and I see some serious issues for the nuclear industry... purely financial ones."
As opposed to those facing competing forms, frankly it's a better bet. We're currently, in the UK, suubsidising offshore wind to the tune of double the market value of the power it generates. We're paying householders 10X the market price of power to use PV. And so on.
"One aspect of public solar generation that is often ignored is the losses associated with the transmission of electricity and the value to the market. About 8% of electricity "
Not here in the UK - transmission losses (as opposed to local distribution and what's politely termed "abstraction and losses" - theft and billing errors to the rest of us) are about 3%.
And you forget something - renewables, almost by definition require long transmission lines. Much more so that conventional or nuclear plant. for example (again in the UK) scottish generated wind power - about half our total potential - has to be transmitted over 400 miles to the heavier areas of demand in the South East. By contrast, Sizewell and Hinkley both have about 60 miles until they're taken off at heavily loaded GSPs, into local networks.
"is lost on it's way to the consumer... so you have to add that to the centralised generation cost."
And it's still a cheaper answer than local, distributed generation. This is a business where economies of scale matter.
I've a small standby genny at home. Running that costs me £0.30-0.50 per KWh - depending on assumptions about how long I run it for (and how I write of the capital). Compare that to British Energy's average contract price at station gate of £0.025.
"Also add that the consumer pays the market price, not the generation price. Solar energy becomes commercially viable for the consumer long before it becomes commercially viable for industry."
That rather depends if the customer wants to stay connected to the grid. See the comments above - solar power doesn't generate a lot at night, and unless you assume a local battery pack - not a cheap, or especially reliable solution, of course - you still need a grid connection.
And, since grids largely are a "fixed cost" business, if you cut the amount of power moving over them, you don't cut the cost to operate - you just increase the cost per unit transmitted/distributed.
"Then there is the immense redundancy of small scale solar generation compared to 6 generators producing 3% of Japan's electricity in one spot."
For redundancy, read surplus capacity - which takes us back to the capacity factor point.
You've missed something, too - distributed generation makes maintenance of stabiulity even on a local distribution grid a big of a problem. One ultra basic example. Say I connect 10KW of solar capacity on the roof of Dawson Towers. And it generates sometimes at that. I've immediately created a power-factor problem (one of the three phases carrying more power than the others), unless by some strange coincidence all my neighbours have done the same.
That way lies serious local instability, and worse, losses on the system. the 3% we see on the UK Transmision grid would look trivial. There's work been done by the 14 PES distributors here, which analysse the impact - they've used it as a case to OFGEM to increase their investment allowances. Even 10-15% of households connecting microgen requires a fundamental rebuilding of the local grid, requiring intelligent substations able to manage frequencies at just about stree level - as opposed to one every 5-10,000 households as at present. It'd cost literally in the 10s of billions.
"Now consumer solar power generation obviously isn't popular with the energy industry"
Well, a couple of my clients are subsidising customers to put it in, and to split the Feed in Tariff benefit, so they don't seem over reluctant - and they're UK "big six" retailers.
I get the idea, you're something of an enthusiast, as opposed to an analyst, btw....
"so they'll play the, perfectly reasonable, 'what about nighttime use'... but shale gas exploitation has just knocked gas prices on the head for perhaps as long as 100 years""
And compare the options - using gas for baseload, as opposed to using it for peaking. Gas and nuclear actually mesh rather well, perhaps better than gas an intermittent renewables.
As it is, I'm a great enthusiast for shale gas development - not necessarily for power generation, instead because it'll have to largely displace oil in transport fuels (either direct, or via "gas to liquids" technology. The implication there, of course is that oil and gas prices will gradually become coupled at the "per KWh" level, as substitution increases. I sepnd about 4-5 weeks a year in India, and it's already usual there for buses and autorickshaws to be CNG fuelled.
Then think - gas usage is about half of oil usage (on a megajoule level). Even subsititing for half of oil production implies a doubling of demand. And the difference is, for a gas plant, fuel cost is 80-90% of the cost of a unit of output.
"I wouldn't bet on it winning more than a few bids... just for access to weapon grade plutonium."
Well, you'd be on to a loser there. Commercial reactors are lousy makers of eweapons grade plutonium.
Recall the way by which it's claimed to be known that the Pu finds at Fukushima are not from atmospheric bomb tests? The isotpe ratios, with a large Pu 238 content? Pu 238 isn't a good idea when you're making bombs. It makes the Pu extremely hard to handle in the production process - it's about 300 times as radioactive as the "useful" Pus.
That arises from fuel having a long dwell time in the reactor - it arises from a gradual build-up of neptunium, itself a result of running fuel to high burn-ups.
If you want bomb plute, it has to be on a very short refuelling cycle - the precise opposite of what you design for on a power reactor.
"better ability to come in on time and on budget"
French experience, and Japanese tells a different story. First build of a new model - Like Olilkuoto - tend to come in badlly late and over budget. By unit 5 or six, they're pretty much to time and cost (as set for the original).
Now it's true that we in the UK liked to build every plant as a prototype - never more than two identical. But that's not the way sensible programmes do it. Curiously, the Germans were nearly as bad as we were.....