Climate change

October 25, 2010

SNE 2060 – are uranium resources sufficient?

Filed under: Uncategorized — buildeco @ 10:42 am

Barry Brook

In the previous SNE2060 post, I considered four possible scenarios for expansion of ‘Generation III’ thermal nuclear power reactors for the period 2010 to 2060. I attached no probability to them, but obviously not all are equally plausible. For instance, I strongly doubt that the TR2 scenario, which followed the WNA high scenario to 2030 (of 1,350 GW) and then continued this onwards to a massive 10,000 GW of installed worldwide nuclear capacity by 2060, will come to pass – at least not using only current-generation thermal reactor technology based on an open fuel cycle. Indeed, the scenario I think to be most likely is TR1, because it fits logically with a synergistic expansion of closed-fuel cycle ‘Generation IV’ technology like the Integral Fast Reactor and/or Liquid Fluoride Thorium Reactor.

I will explore some of those complementary multi-tech pathways (i.e., the Gen III/IV mix) in later SNE2060 modelling. But first (in the next three SNE2060 posts), I want to examine some of the key assumptions and outputs of the open-fuel-cycle ‘Generation III’-only route, with a critical eye. These include: (i) uranium resources (this post); (ii) spent nuclear fuel storage requirements, and (iii) implied build rates of reactors.

Here I’ll consider uranium (U) supply under a situation of no used-fuel recycling (i.e. once-through). (Reprocessing light water reactor fuel rods to create MOX [mixed-oxide fuel] still amounts to using the uranium resource inefficiently, increasing the energy yield from from 0.7 to just over 1 per cent. It is also expensive and does not noticeably help in decreasing the radioactive life of the waste.) I should note also that I’m hardly the first person to blog about uranium resources (e.g. read here and here, as well as the comprehensive assessment given by the WNA here). But I’ll give my own spin on it anyway, so as to keep the SNE 2060 series more-or-less self contained.

The world’s reserves of uranium are currently estimated at 4.5 million tonnes (t) extractable at less than $US80 a kilogram (incidentally, the market spot price for U on 14 Oct 2010 was $US 106/kg). By ‘reserves’ I actually mean ‘reasonably assured resources‘ (RAR, which is typically defined as the mineral resource that occurs in known deposits of delineated size, grade and configuration such that the quantities could be recovered within the given production cost ranges with currently proven mining and processing technology) plus inferred resources (based on direct geological evidence and extensions of well-explored deposits). World production was 50,722 tonnes in 2009, and has grown at an average rate of 7.1% over the last 3 years. World demand from thermal reactors is greater than this mined figure, at about 70,000 t/a, with the difference made up relatively inexpensive secondary supplies (stockpiles, weapons etc.).

Two useful things to note here. First, look at this chart:

See how most identified U is recoverable at <$80/kg, and relatively little more at <$130/kg? This is because it’s not really worth exploring for the more difficult-to-extract stuff until the easy and cheap (and already identified) ores have played out. Once you’ve identified 20 to 30 years of assured supply, what’s the urgency, after all, in identifying decades more? The fallacy of ‘peak metals‘ fails to consider this little economic principle (this is the main reason why Julian Simon won the famous wager with Paul Ehrlich).

But that said, with the prospects of a nuclear renaissance, exploration has notched up over the least few years, as illustrated in this figure:

Note the positive correlation between exploration dollars spent and the expanding size of the identified RAR for uranium. This suggests that with more exploration, there is still a lot more (relatively cheap) uranium left to find. More on these economic geology fundamentals here. It’s difficult to speculate on how much further the RAR for uranium will rise over the next few decades, but it’s likely to be considerable (my guesstimate — quote it back to me in a decade hence — is that it will have risen to ~10 million tU recoverable at <$US 100/kg by 2020).

From a greenhouse gas emissions perspective, Lenzen (2009, pg 49-50) conservatively estimates that if ALL the RAR ore grades between 20% and 0.01% were used to generate power using once-through thermal reactors, the resultant 210,800 TWh of electricity would generate around 17 Gt CO2e, at an average emissions intensity of 80 kg CO2e/MWh. (This is working under the assumption that all the energy used for mining, enrichment etc. comes mostly from coal-fired power stations and oil-burning machines, rather than nuclear electricity and synfuels — parse the logic on that if you are so inclined). For the ore grades greater than ~0.05% U (which is still the majority of the currently identified RAR), the emissions intensity drops to <20 kg CO2e/MWh (Weisser 2008).

Beyond the immediately mineable resource, there are another ~35 million tU in lower-quality ores and mineral compounds (phosphates), which the Organisation for Economic Co-operation and Development (OECD) says could be economically extracted for a few hundred dollars a kilogram (cited in IPCC 2007 Working Group III report, chapter 4). Interestingly, we’re probably going to want to mine these phosphate ores at some point anyway for reasons other than energy, to keep the world fertiliser markets supplied. Indeed, according to a co-authored study by leading American expert John P. Holdren (now Director of Science and Technology Policy for the US) and IPCC Chair Rajendra K. Pachauri, there is ultimately likely to be between 100 and 300 million tU accessible via mining at a price of <$US 350 (in 1992 dollar terms).

Then, of course, there is the huge potential ‘stockpile’ dissolved in the oceans, which contain 4.5 billion tonnes (Gt) of uranium. Even today, we have the technology to extract uranium from seawater  (see figure on left) at a cost of between three hundred to a few thousand dollar per kilogram. If you think this is crazy, do yourself a favour and read the assessment of this possibility by Nobel Laureate Georges Charpak in his book “Megwatts and Megatons” (Pg 210-212, 220; Google Books Extract Here). Now as Mackay (quite realistically) points out, even with refined and scaled up extraction methods, we might only be able to access a tenth of this – say 500 Mt – per 1000 years, due to the slow pace of ocean circulation and the relative diffuseness of dissolved uranium in seawater.

Is this cost-effective? Well, for current light water reactor technology, the total fuelling costs – including mining, milling, enrichment and fuel-rod fabrication – is around $US20 million a gigawatt a year, for “yellow cake” (uranium oxide: U3O8) at a price of $US70 a kilogram. In unit cost terms, that works out at 0.2 cents a kilowatt hour for the uranium “fuel”. So even if uranium ore prices rose by 10 times (to $US 700/kg), it would lift the price of nuclear-powered electricity by 2 cents a kilowatt hour for Generation III tech. This is equivalent to the average household’s electricity bill rising by $150 a year in Australia (at present it is about $1,000 a year). (By comparison, a 10-fold rise in the price of coal would increase annual household electricity costs by $900, while for natural gas, such a price rise would cost $1,700 extra each year.)

Finally, I should link this back to the conclusions of SNE 2060 Thermal Reactor scenarios. The TR1 option required the consumption of about 14 million tU by 2060, and TR2 required about twice that. Clearly, both figures are higher than the current RAR for mined U, but then again, are lower than the U resource that the IPCC estimated was worth extracting from phosphate minerals, and is a tiny fraction of what is conceivably accessible in sea water at a reasonable price. I conclude, as the recent MIT 2010 report did, that uranium supplies are not a serious constraint on nuclear power deployment during the 21st century, even at very ambitious growth scenarios. HOWEVER, that does not mean that I’m arguing that TR2 is the most desirable — or indeed most likely — scenario. Far from it, as I plan to elaborate in later SNE2060 posts, when I discuss the Gen IV synergy. I just wanted to squash a few ‘mythconceptions’, that’s all.


Leave a Comment »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Create a free website or blog at

%d bloggers like this: