Posts by Andydaws

New nuke could POWER WORLD UNTIL 2083

Re: @ Andydaws

"Is the energy density of the salt sufficiently high that decay heat is only a limited meltdown/containment failure risk? So passive cooling arrangements and heat sinks are sufficient."

I've not considered density per se - I don't think it's a major driver in this. What I'm working on is a simple assumption that the decay heat removal requirements in toto are roughly similar between thermal designs - which has to be agood assumption. For a 500MWe unit, even assuming a 40-50% thermal efficiency, you're still producing 1-1.2GW at full power. And decay curves give us 7% or so of full power heat production on shutdown ((70MW or so) decaying over the first day to 20MW or so, and down to 10MW in another couple of days.

That's a lot of heat to lose - and without making the drain down tank huge, or of a geometry that'd be hideous to shield, I really can't see passive air cooling doing the job.

There's also a bit of a misunderstanding about the Oak Ridge reactors. They were extremely small - 5MWTh at full power, if I recall rightly. It's also worth understanding that yes, they were drained down - but also that they operated only intermittently. They also lacked any breeder blanket/salt arrangements, and (I think) had nothing like the full fission product extraction functions that would be needed in a commercial design. And, let's recall - radiological standards in the fifties were by current standards hideously lax!

I agree re the advanced fast concepts - and in all honest, see concepts like IFR as far more developed and ultimately viable that MSR. For the longer term, I suspect that the lead-cooled designs will emerge as leaders - concepts like BREST-1200 using online electrochemical processing of nitrate fuel, and operating at 700--800C have to be attractive. Yes, there are coolant chemistry issues (to do with the behaviour of oxygen in lead), but they look minor compared to the chemical challenges of MSR!

Re: Blimey - Sounds Too Good To Be True - Help Please.

how long have you got....?

basically, it's the same as any other reactor - neutrons arise from fission events, then get absorbed into other fissionable nuclei, which fission and produce more neutrons than they absorb - the idea being you balance the production of neutrons, their absorbtion into fissionable nuclei and other parasitic materials, and losses by leakage out of the core.

The graphite is there because uranium nuclei are more prone to absorbing neutrons moving at around 2200m/s (thermal energies) than they are ones that have just been produced from fission at much higher speeds ("fast" energies). The neutrons bounce around hitting carbon atoms in the graphite (which doesn't absorb them) losing a bit of speed with each collision until they've slowed to thermal levels f energy, and will then tend to be absorbed and cause fission.

Re: Looks like an update to me, not new

I think you're misunderstanding where Wigner energy comes from. It's not heat storage per se - at least the input energy isn't heat, albeit it manifests itself as heat on release.

It's actually energy that's arisen from neutron collisions with the carbon atoms of the graphite crystalline matrix - basically, atoms that have been "knocked out of place" and have re-bonded in distorted patterns. That being the "annealing" point - it's when temperatures rise high enough that those distorted bonds are broken, and the matrix "snaps back" into it's lowest energy alignment state.

SO, unless, you happen to have an intense neutron source to hand, and a material that allows various forms of cross-bonding, wigner ain't going to happen.....

Re: As a few have said, there's nothing much new here....

I've been seeing stuff like that for a good few years, John Smith, and as I said, it's more than a touch naive - and given that we're talking of at least three or four entirely separate physical processes (a spay/sparge for volatiles, vaccum distillation for medium-weight semi volatives, pyrolitics or similar for actinides, and a protactinium extraction process), I think someone's indulging in wishful thinking, if they think a single compact plant will do it.

We've then got minor issues like high-integrity, actively cooled shielded storage for the first two (and probably the space where the protactinium will have to dwell for a couple of years while it undergoes decay to 233U).

And no, simple redundancy won't do it - we're talking highly integrated plant that will be working in direct contact with a medium that's not only laced with fission products, but is in intimate contact with neutron emitters. It'll therefore be subject to significant activation. That's going to mean maintenance access will be essentially zero, unless you allow significant periods to allow activity to decay, before attempting access; which in it's turn means you're not talking about simple duplicate or triplicate redundancy, but manifold. If yuo've ever seen the maintenance access problems on plants like those at Sellafield - which only handle fuel, recall, after a decade or two of decay time, where as this is needing to cycle through the entire fuel load every month or so - you'd not dismiss this quite so casually.

"The design of equipment using molten salt is specialized but not uncommon. ",

However, not at nuclear standards or integrity/reliability. And not dealing with activated products in real time.

"Because the fuel is a liquid the moderator elements are much simpler (essentially rectangular pillars in the referenced report outline)."

To put it delicately - balls. The AGR moderator "bricks" were not notably complex in shape (basically octagonal section blocks with keyways for jointing.. And no, they weren't subject to direct loading from high velocity gas - each fuel channel had an inner liner, and the fuel stringers sit in their own sleeve. Geometry wasn't especially the issue - the reason they tended to distort was that partly there is obviously a neutron flux gradient across the brick. Now, if you think of the scale of an AGR core, and the scale of an MSR core - the latter being very much smaller - I'd reasonably expect gradients to be steeper in a smaller core.

And, I don't see concepts of in-situ replacement being even mildly credible - not after direct contact with fission-product bearing fuel salt. There were some thoughts given to trying that on AGRs in a depressurised state (and they don't have anything like the direct contamination issues) but it was quickly kyboshed as infeasible.

Re: Thorium

No, you need the chemical plant to remove the poisons even if the plant IS run without fuel breeding, or if you run separate fuel/breeder loops.

It should be obvious, really - if you're going to breed 233U from 232Th, it has to involve a neutron absorbtion, then a decay period (the actual process transmutes 232 Th to 233Pa by neutron capture, which then undergoes beta decay to become 233U). The necessary neutron flux is only available in the core, where active fission is going on. Therefore, you're pumping the breeder salt through the core, therefore it's absorbing neutrons, and if you allow the Pa burden to become too high, it closes the reaction down - and you need it to undergo two more captures to become useful fuel, ie 235U). It doesn't matter if it's mixed in with the fuel salt, or a separate circuit, its still absorbing neutrons.

MSRs also depend on active and continual removal of fission products from the fuel loop. Some of that's easy (letting Xe and I come out of solution) and some's hard (CS, Sr and similar via tricks like vacuum distillation, which is a sod, and would be a horror to maintain in a highly active environment). And if there's any production at all of higher transuranics (which there would be in there were any 238U in the fuel salt - and that's inevitably until and if you get to a completely closed 233U based cycle), there's no removal mechanism at all, short of pyrolitics or something like purex.

Re: Even if it does generate Pu

"This is important, when you realise that the high level radioactivity associated with spent rods in current technology is Cobalt and Caesium, with a dangerous lifepsan of about 200-500 years. After that, your waste repository becomes a low-emission, extremely high quality plutonium mine (98% U238) - and before that the Pu is easily separated from the nasties via chemical processes (in a thorium reactor you leave them in solution to burn up. It's only U234 whic h needs to be removed as it will poison the process if levels get too high. It's all pretty elegant, enriching isn't needed and because 100% of the fuel is used, a lump of thorium goes 30-40 times further than the same size lump of reactor-grade uranium (the process can be used to convert u238 to reactor fuel, eliminating 99% of the current volume of "high-level nuclear waste")"

Mostly correct (certainly in the fission products argument), but you miss a few of the "hard to handle" products - mostly the "heavier than plutonium" transuranics like Amerecium. They're the biggest contributor to spent fuel activity after the 200-300 year mark, and will keep fuel a lot hotter than the original ore (the baseline target for disposal) for some thousands of years. And no, an MSR cycle doesn't make significantly smaller proportions of those than a uranium one (after all, it's still ultimately fissioning uranium, with a 238 content in there).

They're fissionable - but not very much so at thermal energies. They can be burned, but only in fast reactors.

That's in large measure the elegance of design like IFR, that use a mixed actinide metal fuel. Using a pyroprocessing method that's basically electroplating in a molten salt medium, you don't ever need to separate the U/Pu/Am etc. from each other. The real virtue, though, is it doesn't have to happen in real time - the fuel can have a couple of years of cooling/decay, to get rid of the really hot stuff, then processed. With an MSR, you have to do it "real time", otherwise the reactor shuts down.

And looking at the reliability challenges of operating plant in a "hot" enviroment, real-timne scares me sh*tless.

Re: connected to a drain plug of salt that has been frozen solid

I think you slightly underestimate the heat removal issue - we're talking in the order of 10-20MW in the immediate aftermath of a fuel drain-down. Expecting to do that simply through convection into air would require a pretty immense large area, nless you assume either pumped flow, or two-phase of some sort - and that implies a secondary coolant supply (probably water)

In reality, it probably two-phase cooling to get the decay heat away - which puts you back in the same sort of place that you've got on the 72-hour passive cooling arrangements on an AP1000 or EU-BWR. That is, a tank above the decay heat source, allowing boiling off the walls of the fuel draw down tank, with water supply by gravity.

Oh, and by the way - that drain-down tank is going to be a pretty huge source of "gamma shine" - which means it'd have to be behind several meters of concrete, and but still have to have good air/water low. And it also means that any equipment associated with it would have to be entirely remote-operation, and maintenance. Once that tank had been used, you'd not be re-entering it's immediate area for a few decades, to fix anything - like the pumps required to move the fuel back into the reactor.

Re: Nothing new here, been squashed since the 60s!

"There some technical issues to solve "

that's something of an understatement.......

Re: Looks like an update to me, not new

"They aren't willing to say if this is thermal, fast or something else."

well, if it's not thermal, there's a very good question as to why there's a large chunk of graphite in there.....

And the graphite energy storage issue (Wigner energy) isn't an issue if the reactor works at about about 300C

Re: Good job - to a degree

The "salt plug" is a useful eature - but people rather miss the point.

The salt plug prevents an ongoing nuclear reaction, by draining the fuel below the moderator, thus making the whole system sub critical.

But that's hardly a unique feature to the MSR, and nor does it remove the key issue in safety - which is removal of decay heat. Note that at Fukushima, in all cases the main chain reaction stopped - it had to as coolant/moderator was boiled out of the core. And in IFR designs, a mixture of doppler broadening and spatial effects also stops reactions as temperatures rise.

Even once the fuel is drained down, it's still necessary to remove decay heat (typically about 7% of full power at the point of shutdown, decaying down to 2-3% within a day. For a 500MWe reactor, that still means you're removing 20MW or so.

In some ways, the MSR makes that problem easier, in others harder. The fission product burden in the fuel salt is lower - because of the "real-time" reprocessing required to make the design work. But the total fission product burden is higher - the xenons and iodines aren't removed by neutron capture - and those and other fission products (roughly the same in quantity) still have to be isolated and cooled. Arguably, that's harder, as concentrating those products means that the cooling needs to be more aggressive (and reliable). Worse, since different products are isolated in different streams, the number of cooling systems proliferates, and reliability engineering 101 tells you that's a bigger challenge, not a lesser.

By th way, if anything 233U is a better bomb material that 239Pu. The claimed MSR advantage is that it's mixed with 232U, which is strngly radioactive - but then, so is 240Pu, which is inherent mixed into reactor grade plutonium. Making a bomb from either is much the same scale of technical challenge.

As a few have said, there's nothing much new here....

But as someone who's actually worked on designing, building and operating nuclear plant (Heysham II, Torness and a dabble in Sizewell B), there's more than a touch of naievity (sp?) in both the engineering and the assumptions about how it could be brought to market. Plus a sonething that looks like some sleight of hand about fuel and waste cycles.

So, in no particular order.

First, MSRs can indeed in theory, be simple beasts, at least in terms of the reactor itself. However, what's usually omitted by the proponents are the issues associated with anciliary plant. An MSR basically demands that you reprocess in "real time" - typically, the entire fuel load has to pass through plant removing fission products and fuel precursors every 2-4 weeks. In the case of some products, that's relatively simple - xenons and iodines can be got out by spraying the fuel salt through an inert-gas chamber. Others - the less volatile medium weight fission products like strontiums and caesiums will need something like vacccum distillation. The most challenging is removing the protactinium fuel precursor (the cycle runs 232Th + n -> 233 Pa - e -> 233U). If not removed, the protactinium captures extra neutrons, needing to capture 2 in order to transmute to 235U. The protactinium has a big capture cross section, and if left in will badly reduce neutron economy, so it has to come out, both for the sake of breeding and basic operation. It's a bugger to remove - the most viable method involves bubbling fuel salt through a column of liquid bismuth.

All of which would be fine - but your working fluid is a highly radioactive molten halide salt at 500-600C. And the plant has to operate at pretty much the same levels of availability as the reactor itsself. designing and operting plant to acheive that is extremely hard - probably harder than desinging the reactor itself.

And don't be fooled by the fact that the fission products are removed from the reactor means that they're not a management challenge. Inventories of Xenons and iodines will be higher than a conventional plant - in conventional plant, they're transmuted away by neutron capture - in this, they're separated, and will have to be isolated, stored and cooled, alongside the other fission products. True, it's a lot smaller than the volume in a cnvetional spent fuel pool - but much hotter!

There are other issues, too. Keeping a graphite core in useable condition in the core of an AGR is a sod of a job - it suffers radiolytic corrosion, and tends to distort under high neutron fluxes (not good, as in extreme circumstances it cound cause control rods to jam). We ended up having to manage flux very actively, and to dope the CO2 coolant with methane to get 30-40 year design life. What this things core graphite will be like, with higher flux levels, and an inherently aggressive coolant/fuel medium doesn't look to be likely to have a long life - I'd guess 10-20 years.

On the fuel cycle, the advocates pull a small trick, in that they compare "once through" LWR cycles with a cycle that entails inherent reprocessing. A more appropriate comparator is something like the IFR/PRISM proposal - which has an on-site pyrolytic reprocessing cycle using mixed actinide metal fuel. The harder neutron flux (i.e. faster) is much better at removing transuranics than MSR's themal spectrum, and is at least as proliferation resistant, and fuel effficient.

Climate watch: 2012 figures confirm global warming still stalled

Aren't both sides missing something?

Always fun to see the more committed members of the pro-and anti-AGW sides beating seven colours out of each other.......

But where does any of this actually leave us (and we'll leave aside the implications of yesterdays announcement about black carbon being a more significant contributor to temperature forcing that had otherwise been held to be the case)?

Well, there's no inconsistency between there having been an extended flat period in temperatures, and the current decade being the warmest in the instrumented record. If you think, unless temperatures were actually falling, you'd expect the latest decade to be the warmest, if there's even a slight underlying upward trend.

And yes, there have been other flat periods in the record, and periods of fast rising temperatures (the 1930s), and even of falling temperatures ((the late '40s and 50s). All of that within a rising trend.

So, we know there's been a rising trend since the 1880s or so.

But, it's not a fast one. It's about 0.7 - 1.0 C/Century.

What was supposed to be important about the 1980s/1990s warming was that it showed that that trend had accelerated significantly, and showed signs of accelerating further.

It's now much more questionable that that's the case. What's more is that the basic forcing models for CO2 alone would give a rate of rise similar to that long term trend that we've seen for the last 130 years.. What was supposed to have changed was feedbacks from water vapour and methane emissions would act to make the rise even faster.

However, neither of those - either water vapour content in the atmosphere or methane levels - actually seem to be rising at any significant rate. So what was always some fairly flaky modelling around those looks as though it was wrong.

To sum it up - things are warming - but it's looking as though there's no real evidence that we've really departed from that 0.7 - 1.0C/century. The more lurid claims of 4-6C by 2100 start to look really exaggerated.

Which puts us in an interesting place - is a degree or so of warming by 2100 really a rate of warming that we can't live with? After all, it's about what we adapted to over the last century or so, without serious problems - and it may even be beneficial in terms of agriculture.

That's what needs to be being discussed.

Hitachi buys Horizon to save UK's nuclear future

Re: @Mike Richards - PWR, CANDU and BWR

CANDU steam conditions are if anything slightly worse than those of PWRs - broadly similar to those of a BWR. And, of course, they're not the easiest to build - lots of on-site fabrication to connect up all those little pressure tubes.

On plutonium, the LAST thing you'd do, if you were after bomb grade Pu was got for a design with a long refuelling cycle/high fuel burn-up. That's when you get a build-up of Pu240 and higher actinides, which are a very bad thing when you're fabricating a bomb. So, if anything, CANDU types (no enrichment, on-line refuellable) are a much better option for bomb making than LWRs of any sort.

One other thought on an LWR - this from someone who worked on AGRs - they're one very large inherent advantage. Lose coolant, you lose moderation, and you're inherently in shutdown. Suffer a pressure vessel brach in an AGR, or a major pipe break in a CANDU, you can at least in theory have a full scale reaction ongoing, and no cooling (in CANDU, at least until you get voiding in the calandria).

I've no idea how CANDUs are secured against this - on AGR, there's a "s**t or bust" option to dump a ton or two of boron-glass beads into the core. It's a system tht gives operators nightmares. If it ever were to operate spuriously, it'd write off the reactor in about 2 minutes flat.

New nuclear fuel source would power human race until 5000AD

Re: alternative to nuclear from the oceans

Oh, let's think.

Say I want a water turbine to produce 3200MW, like the EPRs destined for Hinkley Point.

Average tidal current in the favourable sites around the UK is about 2m/s, and there's a fundamental physical limit (Betz's law) which limits the amount of extractable energy to about 65% of that in the incident stream of a fluid.

Energy available (E) = 1/2 mv^3, where m is given by rotor area (A) x density (D), so 1/2 * A * D *v3. Include Betz's law (B) to get

A = 2E / (B * D * v3)

density = 1,000 kg/m3.

So including Betz's law, the best possible turbine needs just under 400 m2 of area - or to be about 22 metres in diameter. That's one f*ck of a propellor. Even those on vessels like the "Ronald Reagan" are less than 1/3rd of that size.

You also can't p[ut them (obviously) in water that's too shallow (which is a pity, because that speeds up tidal flows). Or too close together, because you get flow disturbances in their wake. So, to allow for replacing Hinkley Point C, (and adjusting for capacity factors), you'd need something over 4,000 of the massive devices.

Mounted offshore in water at least (say) 40-50 metres deep.

Re: Time will tell

You don't "drag" them. You just suspend them in a current.

Re: we know that nuclear power is safe

"If you only care about body-count, then nuclear scores pretty well. However Chernobyl and Fukushima have left large areas of land uninhabitable for the next 1000 years or so."

Background radiation levels in Pripyat area about 1 microsievert/hour - or about 9 millisieverts/year.

The average radaition level across Cornwall is about 8 millisieverts/year, with hotspots up to about 6 times that.

Should we be evacuating Cornwall?

Greens wage war on clean low-carbon renewable energy

Re: Fusion might work someday. Fission works *now*.

I am one of those engineers, at least by original training (and has found the standards of the private sector "British Energy" markedly higher than they were in my days when I trained with the CEGB/NNC, but we'll leave that to one side for the moment).

There's one essential you seem to be missing here.

Cost.

I'll agree that fuson is likely to be one day technically viable. Don't forget however, there is a major radioactivity problem - the majority of the energy leaves the fusion reaction in the form of high energy neutrons, which not only activate the reactor structures themselves, but require the reactor to be surreounded with a capture blanket, in which these neutrons can first transfer their energy to a moderator type material (probably lithium or similar) and ultimately be captured. That too is likely to involve significant production of radioactive products of various types.

However, we're seeing nothing at the moment that suggests that fuson plant is likely to be other than horrendously expensive - ITER, even at just 500MWth is looking likely to cost well north of €20Bn to complete; that's for a plant that lacks any steam-raising or other "balance of plant", and would be good for perhaps 150-200MW of electrical output. Weas yet don't have the faintest idea of the likely operating costs of a commercial scale plant, beyond knowing that they're likely to be extremely high - those neutron flux issues will mean that almost everthing will require remote handling, and the plant is hugely more complex than the internals or associated plant of a fission reactor.

Even assuming we we able to add all the balance of plant without inflating costs, and reduce by a factor of ten the unit cost of a fusion plant itself (on a €/MW basis, by scaling and avoiding FOAK costs), and achieve similar unit operational costs to a fission plant, we'd be looking at unit output cost in the £350-400/MWh range - or about 5 times the unit cost of output from a new generation fission plant.

That's a phenpmenally tall order.

Reg review of 2011: Jobs, floaters and 90,000 tons of radioactive water

whicwould hardly be an issue

for soluble material like casesium, would it?

I despair of people's understanding of the most basic chemistry and physics.....we obviously have an educational system mainly designed to produce shaved monkeys.

UK nuclear: Walking into darkness with eyes screwed shut

This has to be taking the p*ss

Let's get this clear...you think I could connect a 1000KW motor to a suitably sized generator and get out MORE than 1000KW from the generator?

Here's a funny thing - that's exactly how the shore stations for the UK=France electricity interconnector used to be set up (anyone guess why it was needed?)

That sure as hell doesn't get more power out than goes in!

"he's a busy chap, even when he's not being a professor"

He would be - he does 4 days out of 5 at DECC, which is where I came across him.

You need to do a few sums, or understand a bit more about the power system.

If EV numbers were anywhere near sufficient to make the scheme worthwhile, there'd be no "off peak" to speak of - in fact, short of rewiring the distribution grids, you'd be limiting charge rate to a prevent overloading at night, rather than in the day. It's one of the few areas where smart meters look even marginally worthwhile.

"Did you see the bit where I said "school/supermarket run"? How far away is the typical school/supermarket?"

Well, that rather then depends on assumptions about what proportion of vehicles are used that way - which probably takes your available capacity down by rather a lot. We've something like 70% female participation in the labour force, so most people don't just need the school-supermarket run - they need a more useful range than that.

There's very good reason to think they'd have worked

Since all but one of the gennies fired up and ran, until they lost aspiration and/or fuel due to the Tsunami.....

Here's the most detailed timeline analysis of what happened at Fukushima

http://www.nei.org/resourcesandstats/documentlibrary/safetyandsecurity/reports/special-report-on-the-nuclear-accident-at-the-fukushima-daiichi-nuclear-power-station

download the document, and have a look for yourself. Ironically, the only one that didn't fire up was because it was down for scheduled maintenance.

Incidentally, I've just had a look around for anything claiming the gennies weren't maintained. I can't find anything. Do you have a source, or is it "something you heard in a pub"?

Ironically,

the only environmental regulation that would prevent that would be that the dumping of waste heat would raise the temperature of the Thames more than would be permitted.....

A few generations is all that's needed.

The overwhelming majority of the radioactivity comes from fission products, not the uranium or other actinides. And anyway, the latter is useful as fuel.

Separate out the fission products - the strontiums, caesiums and so on - and you've got stuff that's broadly got half-lives between 2 and 50 years. Ten half-lives reduces radioactivity by a factor of 1000 - or, would take a fission product mix down to about the same level of activity that the uranium originally came from.

Molten salt reactors using thorium

If that's what you've been told, you're much misled.

Xenon doesn't accumulate in conventional fuel - it's removed by neutron capture (that's the point of it being a poison), and reaches an equilibrium level within a few days of starting operation.

Conventional reactors aren't refuelled because of Xenon - they're refuelled because there's a progressive reduction in reactivity (as U235 is burned), and when other fission products accumulate - typically stuff with atomic weights in the 90-130 range - things like the strontiums and caesiums.

Those are still produced in U233/thorium reactors (note, the fuel isn't thorium - thorium is a breeder material which is converted into U233 which then powers the reactor). They need to be removed.

And that's where things start to get hard. In a coventional reactor, you start off by using enriched fuel. When the fuel is insufficiently reactive, you remove it, then store it for a couple of decades. Then, if you're inclined, you reprocess it. By that time, 98% of the radioactivity has decayed, as compared to when it first comes out of the reactor.

In the MSR, you have to get this stuff out in "real time". You need the processing plant to run at the same or higher level of availability as the reactor itself. You then need the ability to manage and store this freshly extracted material while keeping it cool. The MSR protagonists usually suggest a vaccum distillation process.

Then, if you want to fuel the reactor on Thorium, you've also got to be breeding. That involves a process of neutron capture trom Thorium 232 to Protactinium 233, which then decays to U233. You need to get it out quickly, because it's not only a strong neutron absorber which would screw the performance of the reactor, but if it captures a neutron, it's then got to be left in for two more neutron captures until it can produce U235.

The only way that's identified, so far, to do that is to bubble the thorium/protactinium salt through molten bismuth. That in it's turn has the delightful property of activatiing to Polonium-210 - the stuff that was used to poison Aleksander Litvinenko!

And you're doing all of this (and other things, like sparging out the Xenon) while handling a corrosive halide molten salt at 600-700C - and, should the flow be interrupted, and the salt goes solid, you've all sorts of problems.

It means you've got to be able to build and run a complex chemical plant, dealing with some hideous material, in a highly radioactive environment (which buggers most ideas of repair and maintenance), at the same level of reliability as the reactor itself.

I'm not about to say it can't be done - but anyone who thinks that's going to be done with simple and cheap engineering, or that the design and maintenance challenges are going to be any less than those of a conventional plant is fooling themselves.

Which is what happens with first of a kind complex engineering projects

The B787 was announced in 2003. It was due a first flight in 2007 - it actually happened in 2009.

The Airbus A380 programme was started in 2000, for service in 2005. It was in 2007 - and instead of the development costing €11bn instead of the budgeted €8Bn.

Does that mean that every future B787 or A380 will be delayed in construction, and over-budget?

Sorry, where does this $30Bn number come from?

Let's be clear about what the Chinese are doing.

The origin of this story seems to be something in "Wired" magazine back in FEbruary. It reported a speech at the Chinese National Academy of Sciences saying that they'd do some planning and lab-level studies on Thorium systems. It didn't mention any sum at all, much less something like $30Bn.

That seems to be an internet myth.

What they're actually committed to building is about 70 - count them, 70 - conventional reactors. They'll be a mixture of derivatives of the old Framatome PWR design, and an uprated 1400MW version of the Westinghouse AP1000. Conservatively that's $100Bn or so. They're investigating further upgrades to 1700 or even 2100MW, if the passive cooling capabilities can be scaled up.

They've signed deals to build sodium cooled fast reactors, based on the Russian BN-300. That's probably about $10Bn.

They're building a near-commercial scale HTR - gas/graphite - using a derivative of the old pebble-bed concept; that's probably a $billion or two.

The reality is, the thorium stuff is at best a sideshow, to a programme that's going flat out on conventional nuclear technologies.

At the risk of stating the bleedin' obvious

Chernobyl took place in a "not for profit" electricity system. And the Windscale fire wasn't exactly on a commercial plant.

You presumably don't mind climbing aboard a commercially operated aircraft, do you?

Ultimately, it's not the plant operator who decides the standards under which the plant was licensed. It wasn't TEPCO that decide an eight metre sea wall was adequate - it was NISA, the Japanese regulatory agency.

When, many years ago, I was modelling the risks from various incidents on the Heysham and torness plants, it wasn't the CEGB or SSEB who decided what constituted the 1,000 year storm against which we had to qualify - it was the Nuclear Installation Inspectorate.

And then, in the morning,

come out and find you've only a few miles range in your EV.

I have told Mackay, too. He didn't have an answer.

A few numbers, if you want.

So far, eight light water reactors of the types proposed for UK new build have been comprehensively decommissioned - including the original Shippingport PWR, Trojan (a near 1,000 MW unit), Haddam Neck (BWR) and others. In these cases the entire reactor has been removed, the containment and ancialliary buildings removed and the site released for unlimited use. In some cases, however, some dry-cask storage for spent fuel remains on site. The fuel should have been removed to Yucca Mountain, but the political paralysis around the repository means it remains on the original site.

Decomissioning costs have been in the range $600-1000 per KW of capacity (the smaller plants tend to cost more). Note LWRs are much cheaper to decomission than the gas cooled designs we built, not least because there's not a 2,000 tonne graphite core to dismantle and dispose of.

We'll take the mid-point - $800/KW - and not make any assumptions about learning curves, or the fact that new designs are "decommissioning friendly".

An EPR is 1600MW - so, £800M, if we assume a $1.6:£1 exchange rate. Assuming the EPR hits the same through life capacity factor as Sizewell B's managed so far - about 88% - it's make something like 740 million megwatt-hours over it's 60 year design life.

That's about £1.08/MWh. A MWh currently trades (wholesale) at about £50.

If you assume 2% interest rates (i.e. German Bonds), you actually need to set aside about 1/4 that amount.

Google gives up on saving world from cheap coal energy

There's a bit of an issue about putting solar thermal systems in deserts....

You need a heat-sink - somewhere to dump the heat once the steam (or whatever) has been through the turbine.

It's usual to use either the sea, or (via cooling towers) the atmosphere. Even if it's the latter, it still needs lots of water; even reasonably efficient stations evaporate 5-6 million tonnes/year for each 1000MW of capacity.

There doesn't tend to be a lot of water in deserts - that's why solar development has been banned across most of the Mojave desert in the US, for example.

Desertec isn't PV - it's concentrating solar thermal....

And if PV's so good, what do you use at night?

Greens threaten to sue over solar power cash slash

Renewables impact? = minimal

It's true that emissions have fallen - but it's almost entirely driven by the ongoing switch from coal to gas.

the "dash for gas" took place through the 1990s - we went from a near-zero gas contribution to about 40-50% of production. It's only in the last year or two that there's been a small (1-2%) shift back to coal as gas has been expensive at times of peak demand. And even though the nuclear fleet has shrunk, it's tended to be run harder/longer.

The difference in CO2 output/Kwh for coal and gas is huge - compared to a 1960s-era coal fired plant, an modern CCGT unit will put out 40-50% of the emissions per unit output.

Solar power boom 'unsustainable', says Gov

No, they won't

"So in total all the solar installations generate the same amount as a decent power station?"

Don't mix up "capacity" with actual output - you have to multiply by the "capacity factor".

In Germany, who've had this scheme running for a decade, they see an average capacity factor of just 9%. So, the 327MW quoted is likely to produce about 29MW.

For comparison, EDF yesterday submitted its planning application for Hinkley Point C - two reactors making up 3,260MW. Assuming they run much as has Sizewell B so far, that'll average about 2,950MW.

Assuming costs are similar to those in Germany, that 29MW average will have cost about £1.1Bn for kit with a 20-year life. The 2,950MW will cost about £9.5Bn for a plant with a 60 year design life.

And also set it in the contest of our average and peak demand - 40,000MW and 60,000MW respectively.

I'd love to see those numbers

Because looking at the numbers for new build - where there's now good data from the 8-10 light water reactors that are well advanced in decommissioning, about half of which have restored the site to unlimited use - even including NPV effects, the cost of decommissioning is 2-3% of the value of electricity generated over life. Add in the NPV effect and it's even less.

Typical decom cost is $500-1000/Kw of capacity (if anything, that should come down as the experience base improves). For a 1600MW EPR that's £1.2Bn

running for design life at 85% capacity factor for design life (60 years), and averaging £60/MWh is £42Bn. That £60Mwh provides for a return on capital of about 8-10%.

I'll do theNPV version later, but even assuming prompt decom on shutdown I'm pretty certain it'll be under 1%.

Any guesses who was SoS for Energy when this farce was introduced?

A certain E Miliband.....

Energy minister gives grudging nuke endorsement

Not that many are

on interruptible contracts, AC. They're far more common on gas (where end users can have local CNG or LPG storage). Gas interruptions have been exercised in both of the last two winters.

One other comment. I spend a good few weeks each year in India, where (for different reasons) the power grid isn't especially reliable. The reaction of businesses there to interruptions there isn't to shut up shop until the lights come back on - it's to fire up diesel or other standby generators.

With frequent interruptions, you'd see the same behaviour here. Which would be hugely counter-productive from either an energy efficiency, cost or carbon output perspective.

And who's focussing solely on the domestic sector? I've addressed industrial and commercial use, and why demand management isn't about to make over-much headway there (imagine the impact on Corus of being told to cut power in a couple of hours in the middle a pour of steel, or Network Rail in the middle of the London rush-hour). Or finding your local Waitrose shut. Or your ISP going off line.

The electricity market is split broadly half and half residential and Industrial & Commercial, so domestic impact would be just as useful - if it were available.

And yes, e7 and other off peak tariffs are actively marketed. See EDFE's "20:20" campaign as recently as this year, or look on youswitch.

It wasn't addressed to me, but re the gas CHP units you talk about. Think on two things....

First, how well coincident is your electicity and gas usage - or well matched? If I take my own household, we don't run the heating normally between early may, and october. My heating starts in winter at 5:30, but we don't get up and start using significant power for another hour. The heating goes off at about 7:30 because the house will then stay warm until 8:00 when my wife and daughter leave.

The patterns are just as mismatched in the evening.

DECCs work suggests energy use is reduced by just 7% or so - which is more than offset in both carbon and energy security terms because it increases gas dependency. Never mind the cost implications of hanging a complex fuel cell off the side of a simple gas boiler!

David Mackay goes into this at length in "Sustainable Energy without Hot Air". He comes down on the side of a shift to heat pumps, ground source where possible, or air source otherwise (gshps are a bugger to integrate with anything but underfloor heating systems).

AC, If you want chapter and verse,

See below for "operating the transmission system in 2020" from National Grid. It cites evidence on multiple occurences of huge shifts in wind output on that a two hour horizon. (pages 23-27).

http://www.nationalgrid.com/NR/rdonlyres/DF928C19-9210-4629-AB78-BBAA7AD8B89D/47178/Operatingin2020_finalversion0806_final.pdf

As to "interruptible contracts", what industries do you see shutting down at three hours notice? It wouldn't include any process plant (or indeed, most manufacturing). Or shutting down datacentres and so on. In fact, I can't even really see the likes of Tesco being up for that.

For what it's worth, the two largest users of electricity in the country are Corus and Network Rail. Neither of those is about to drop off-grid at short notice.

"Short term" means less than three hours in grid terms, AC - ramp-ups of any plant that involves significant steam generation take at least that long, if you don't want to bugger the plant from things like thermal stressing and fatigue.

Onshore wind

If you look at the "Renewables Obligation" Annual Reports, you'll see that the rate at which onshore wind is being deployed is falling - it was 700MW in 2007-8, 600MW in 2008-9 and 500MW in 2009-10 (the latest year available)

That's because there are huge planning objections to both the farms, and to the transmission infrastructure required.

If you apply the aveage capacity factor of onshore wind to those, you'll get an idea how little actual production can be expected - just 130MW on average from the turbines deployed in 2009-10, for comparison. Or, 1/24th of what'll be made at either Sizewell C or Hinkley C.

And no, I'd not draw too many conclusions about there being a few turbines in France - all politicians do the odd bit of greenwashing.

typo - that's £0.69/MWh.

Still tenfold todays wholesale prices (and excluding transmission costs, etc.

Making a simple assumption the amount of transmission infrastructure would also increase pro-rata, and with typical transmission charges today of about 1.5p/unit, you can add another 18p onto that. So, about 87p/unit.

Which'd probably have you and I paying about 95p/unit, after distribution charges, retailers costs and margin and so on.

That'd put the average domestic power bill up around the £6k mark. If you also assume replacing gas with electricity, even using off-peak power and heat pumps, it'd be about £11k.

And using the usual definition of "energy poverty" - more than 10% of income going on power and gas - it'd mean about 98% of us would be "energy poor"

AC,

not to anything like the same degree, no.

NG (and most other grid operators) run a policy of maintaining spinning reserve or fast-response hydro sufficient to accomodate the largest single potential sudden loss to the system, plus a bit. As this stands that's treated as 1200MW (for Sizewell, or for the loss of 2 out of 3 of the French interconnector supplies).

Grid recommend for wind, that the spinning or short term reserve is equivalent to the average production of the total amount of connected wind capacity.

So, if we were to hit the last government's target of 55GW of wind and maintain current wind capacity factors (a bit over 25%), we'd need about about 12GW of spinning, or short term reserve.

Grid also recommend that there is basically full back-up for wind available within 3-24 hours.

What's been the case for the last few years is that under the last government, DECC opted to accept lower system resilience, so that the back-up costs for wind would be avoided. Instead of doing what ought to be good practice, and running a reserve equal to both the worst case single unit loss plus the short term varaiability of wind, they left Grid's instructions re reserve as it was.

That will change, under the proposed market reforms. A specific "capacity payment" will be brought in to incentivise availability of spinning and long term reserve capacity. However, it's costs won't be loaded on the form of generation requiring the reserve, but "smeared" across the whole market. So, yet another low-profile subsidy for wind.

"So its *average* output is close to *two* usual sized UK power stations."

Well, the EPR is designed at 1600MW, with an inherent stretch to 1800. And there's a design stretch for the AP1000 in the works to 1400MW. So, nearer to one than two. And both are designed around a 60 year design life.

The problem is, cost.

Even the Sustainable Development Commission saw a 1.8 GW average Severn barrage coming in at the £20-25Bn mark. Which conservatively prices the ouput in the £180/MW range, roughly comparable to offshore wind. That's excluding the system costs of the variability, by the way. Colin Gobson, former COO for National Grid estimates the incremental life-cycle costs for offshore wind are increased by about 50% by such afctors (transmission investment, back-up and variability compensation, inefficiencies in the compensating capacity) - that wouldn't be so bad, because there's at least some predictability, but it's still unlikley to add less than 25%.

The last review decided that the barage was unfundable - even assuming large government interventions. I've not seen anything to suggest otherwise since

Demand Management?

No be honbest, I doubt a great deal.

The national Smart Metering trials finished about this time last year, and they're not suggestive that domestic demand is readily influenced - according to the NAO's review, even with self-selecting volunteers in the trial ( as they dryly put it "not nationally representative") the reduction in use was only 2-3% - and even then something like 20-25% dropped out in the year. It showed NO impact in terms of peak shifting, which is what would matter more.

I've seen internal survey work for one of the "Big Six" that examined what sorts of price differential it'd take to get a major shift in usage patters - to displace even 10% of peak demand would need price ratios (as compared to normal day rate) of between 5 and 10:1. Even the in-day wholesale market rarely varies by that much in any 24 hour period.

There's a common sense "sanity check" - the uptake of "economy 7" and the various off-peak incentives. I use it - on my deal the differential between day and night is between 2 1/2 and 3:1. Quite a differential, and with no great effort (timers on the dishwasher and washing machine, and a timed immersion heater running in summer only) I can shift over 1/3rd of my usage to night rate.

And uptake nationally of E7 and similar is under 10%.

That's a much easier sell than getting people to shift from peak - think of it this way. How many people would shift their meal times because the electricity to heat the meal was going to be a couple of pence cheaper if they ate at 9pm instead of 7? Or is my mother about to decide to delay putting on the kettle at 6?

On the industrial side, there are already very heavy incentives to both reduce usage and shift. Most really large users are on contracts called "flex" where they can pre-reserve power at a fixed price, and anything up to three says in advance opt to trade it instead. They already see the full impact of demand peaks.

So, I don't see a lot of mileage there. The market's already been making it happen, for some time - if you follow a trend of "energy intensity" - i.e. the amount of energy used to produce a unit of GDP - it's been falling for 30-odd years.

One other thing to consider. If you do assume decarbonisation is a worthwhile agenda (and I'm not going into THAT debate here....), a major shift away from direct combustion of fuel domestically, industrially and in transport is a prerequisite. That alone should drive a major shift to electricification. On conservative numbers, even getting halfway to HMG's carbon targets tells us we should be assuming about 75% more electricity usage in 25-30 years time, shifting average demand from 40GW to 70GW or more.

Note the dates

"Stuk requests more details on EPR systems

04 June 2010

The Finnish nuclear regulator has said that it is satisfied with the modifications proposed for the design of the control and safety systems of the Areva EPR under construction at Olkiluoto, but clearer documentation on the independence of the systems must still be provided.....

......The Finnish radiation and safety authority, Stuk, which first raised queries about the EPR's systems in December 2008, has now reviewed technical plans submitted by Teollisuuden Voima Oyj (TVO) concerning the control and safety systems of Olkiluoto 3. It concluded that "no notable change" would foreseeably be needed for the planned design.

However, whilst noting that considerable progress has been made in the design of the systems, Stuk said that it "has requested TVO to update the reviewed documentation in such a manner that it provides the necessary initial information of detailed system design in an unambiguous format."

In particular, Stuk requested that "the principles of securing the mutual independence of systems backing up each other are defined clearly enough.""

In other words, issue closed, now catch up with the paperwork.

You fail to note, by the way, the 3rd and 4th EPRs (at Taishan) are pretty much on time and budget - much as you'd expect with series build.

They're both "Ebb flow" designs , Adrian

If you design for bidirectionality, you reduce the total flows available - worse, you tend to end up with complex multi-reservoir designs.

Here's FoE's comment:

http://www.foe.co.uk/resource/briefings/the_severn_barrage.pdf

Note

"The installed capacity, or maximum output, of the Project proposal would be 8,640 megawatts (MW) or 8.64 gigawatts (GW) and would have a load factor of about 23 %. Generation would occur on the ebb tide."

Soorry, should have added this earlier...

There is a way of getting ahandle on what an "all renewables" grid would cost - at least, as a baseline.

Grid planners use a concept called "firm rating"(or "Capacity credit" depending where you are in the world). National Grid use a definition that basically rates a station dependent on the power it's 90% probable to be able to put out at any given moment. Typically, for a conventional or nuclear station, it's 95-99% of nominal capacity let's call it 97%.

NG rates wind as 8% firm, based on operating experience. That means, for every 1000MW of windpower, you're 90% probable to see 80MW or more.

If we make a simplifying assumption - that wind unit operate truly independently, which isn't true, as we'll come back to that in a moment, we can do a simple sum to calculate how much wind capacity we're need to have the same grid reliability as we have at the moment.

We run about 70GW typically declared available to the Grid, excluding renewables. We'll make another simplying assumption, that that's in 1000MW lumps. At the simplest level, we'd need (70*97/8) = 849GW of installed wind. to match that adjusted for firm power. Average demand, btw, is about 40GW.

That's obviously not feasible onshore. so, let's assume most of it ends up offshore. Even then, most would have to be far offshore - so let's take a nice convenient inshore example as about representative - like the London Array.

That's costing about €1.9 for 640MW of nominal capacity. Converting that to £, that's roughly £2.6/GW.

The lowest cost of capital for a significant European generator is EDF - 7.8%. Let's assume they'd gear, and could borrow at just 5%.

849*2.6*0.05 = £110 Billion per year in financing costs alone - we've made no contribution to amortising (paying down the plant), O&M or whatever. To also get the plant to pay for itself over a 20 year life you need to bump that up to about 9% Add another 2% on for O&M (which is well under half the prevailing rate for O&M on offshore structures in the North Sea in oil), and you're at 11% to simply break even.

So, we'd need to recover about £242 billion per year

My net demand will stay the same - 40GW average. That's 350 million megawatt-hours/year.

I make that about £690/MWh - or £6.90/KWh. Average wholesale at the moment is 5-6p/KWh.

There's an underpinning error, by the way, in that - it wouldn't give matching reliability to the current system, because you've still dealing with coupled weather systems. NG reports (Operating the Transmission System in 2020) that it observes several multi-hour period per year when wind output UK wide drops close to zero - often associated with peak demand.

So, you'd pay more than ten times as much for a less reliable system. Sounds like a good move!

You need to understand a bit of the history, AC

Energy was the model followed in when the local loop unbunding came in - the transmission infrastructure was separated from generation, retail and distribution at privatisation, and separation of distribution from retail was enforced from the mid nineties (first on a "chinese wall" basis, and now almost total separation of ownership - the only part of the network in England and Wales owned by a firm that's also active in retail and generation is the old Manweb patch)

And we did indeed run a "pool" - compulsory short term auctions - from privatisation to 2001. It didn't work especially well, being prone to "gaming" by generators, and causing a lot of instability in price.

For those reasons, it was replaced in 2001 by something called NETA - "New Energy Trading Arrangements". You can judge the effectiveness of it by the fact there was something like a 25% drop in average wholesale prices in the 12 months following it's introduction - and prices (allowing for gas and coal prices, which are internationally traded) have if anything fallen further.

It's also worth saying that both OFGEM and the Competition Commission have been over the generation trading arrangements several times since, and found no evidence of collusion to drive prices. And although its true most power comes from generating capacity owned by the big 6 there's significant capacity that doesn't, or has to be traded outside the parent group. The two biggest examples are Drax (4,000 MW), and Eggborough (2000MW); plus as part of the deal to buy British Energy, EDF signed a deal to sell half its production outside the group (something like 5,000MW. Also, the production from the remaining Magnox stations - about another 1500MW - is not part of any of the big 6. Add that lot together - I make it about 12,500MW - and you've something approaching 30% of total generation. More than enough to establish a reasonably liquid market, and to get a price baseline.

Part of the problem is, the press is lazy - they conflate concentration in retail with concentration in generation, and don't look at the internal structures of the market. For example, Centrica has no choice but to trade for 75% of its electricity requirements (and they're the biggest player - 25% of the residential market). They simply don't possess enough generation, so buy from just about everyone out there. By contrast, EDF/BE and SSE are notably long in generation compared to retail - EDF/BE produces about 25% of total production, but only has about 10-15% of the retail market. Indeed, as I commented elsewhere, SSE's gambit about open auctions is more to do with the fact they're short on sales for their generation, and their plant's running horrifically low capacity factors as a result.

Not quite, Bluenose.

Say you pay £10 for electricity at the moment. About 70% of that is the wholesale cost of the power at "station gate" (of which about, with out current generation mix about 45-50% is fuel, 15% operational and finance costs, 6-7% is renewables subsidy, and the rest generator's profit margin...and yes, those are meant to add up that way) . Transmission and distribution together add another 15-20%. Metering and the operations cost for the retailer is about another 5%.

Retailers gross (not net) margin will hit about 9% this quarter before declining to about 6% next year. For the last couple of quarters, they've been around the 1% mark. Which is below the reatiler's cost of capital. Averaged over the last ten years, margins have been about 1-2%.

There's a more cynical explanation for SSE's position

If you look at their website, you'll see they've got an extraordinarily low capacity factor on both their thermal plant and their renewables over the last three months - 23% and 13% respectively.

I think they're having trouble absorbing their power generation across their customer base - and they've been locked out of contract for independent retailers because EDF committed to sell half of British Energy's output to non-EDF retailers.

Winning new UK pylon design may never be used

All quite right

but its actually even worse - for waterproofing, buried or trenched cable needs to be gas-filled and pressurised.

Now, that gets REALLY expensive...

At the risk of stating the bleedin' obvious

How big is the Olympic park, versus running a line from the Scottish border to London.......?