• Increase font size
  • Default font size
  • Decrease font size

Apr 2007 - Uranium Numbers

Those other uranium numbers

At Paydirt’s Uranium Conference held in Adelaide in March 2007, the other actual uranium numbers were covered, such as how much U3O8 yellowcake is used in a reactor, demand/supply and what prices are received. Uranium must be a bit like a mathematicians dream with so many numbers involved. Although asking producers and even nuclear representatives at the conference what percentage U235 is in U3O8 resulted in a surprising range of about 0.71% to 1.0% to don’t know.

So what. Well, most of the Australian producers are using Adelaide Control Engineering (ACE) who have patents over the front and back ends of the treatment process that changes uranium-bearing rock (sand etc) into the U3O8 yellow cake that the uranium mines produce. ACE designed Beverley (about 8 years ago) and more recently Paladin’s Langer Heinrich and Kyalekera.

While a typical plant costs about A$100m to A$200m (depending on its size), the process route is relatively complex due to the number of different uranium minerals and associated gangue minerals. Reputedly the tricky bit is producing the final yellowcake product – a classic light yellow coloured clay, of which the active fissile ingredient is U235 (with a half-life of 700 million years, also written as 235U or uranium-235), with the rest being mainly U238.

The U235 comprises only 0.71% of the U3O8, and is enriched to result in a product for a BWR (Boiled Water) nuclear reactor containing typically 3.5% U235, or in other words 5t of U3O8 becomes 1t of enriched uranium for use in a BWR nuclear reactor capable of producing about 1000MW to 1250MW. A typical BWR nuclear reactor uses about 300tpa of enriched uranium (or ~1,500tpa of U3O8 yellowcake).

PWRs or Pressure Water reactors (also referred to as EPRs or evolutionary power reactors) use 5% U235 enriched uranium, and are regarded by some as the best, being third generation with a life of about 60 years (earlier reactors had a 20 to 40-year life) and generating about 1250MW to 1600MW. They may have a lesser quantity of enriched uranium, but because its enriched uranium fuel has a higher loading (5% vs 3.5%) it may result in higher uranium consumption.

Although the annual “usage” of a BWR is ~300tpa of enriched uranium, it is actually delivered in 2-year consumption lots. Basically, when a nuclear reactor is started, the initial “charge” is about 5 years of annual production and every 18months to 2 years, 2 years’ worth of newly enriched uranium replaces the spent enriched uranium (which is currently planned to be disposed of in special concentric cylinders stored in mine-like workings ~500m below surface).

Necsa (South Africa) proposes to use a PBMR (pebble bed modular reactor) which is much smaller (at 160MW) than conventional reactors, and only requires 50tpa of enriched uranium. However, its enriched uranium needs to be 10% U235, or about 14t of U3O8 becomes 1t of PBMR enriched uranium. As before it needs 5 years initial charge and changing 2years’ worth of its enriched uranium about every 2 years.

According to Areva, there were almost 440 reactors in operation globally in 2006, of which about 360 were either BWR or PWR, 40 were Canadian designed Candu’s and 22 were British gas-cooled (and earmarked for closure). Approximately 129 were in Western Europe, 126 in North America, 109 in S & E Asia, 67 in the CIS, 4 in South America and 2 in Africa. There are a further 23 under construction, plus 39 in the design phase as detailed in Areva’s March 2007 results presentation (available on their website).

With Areva supplying the fuel for almost half of the world’s PWR and BWR reactors and constructing most of the new and maintaining large amounts of the old reactors, they are probably the most accurate of the forecasters of supply and demand. They estimate current reactor demand at about 65,000t of U3O8 (which seems about right for 440 plants averaging possibly ~1500tpa of U3O8) as shown in Figure 1.

Areva estimate that in 2005 they supplied the fuel for about 45% of the world’s PWR and BWR reactors, hence it was interesting for them to comment in their March 2007 presentation that the average selling price to the reactors in 2006 was US$23/lb of U3O8, being only US$24/lb in 2H06, despite the average spot price in 2006 being US$48.5/lb and peaking at US$72/lb at the end of 2006.

Areva commented that sales to reactors are under contract, so the mines supplying the reactors do not receive spot prices. Selling prices were briefly covered in the Paydirt conference in reply to a comment by another company that Paladin only receives US$55/lb for its yellowcake from Langer Heinrich. The reply from a Paladin representative, was that was not correct, while it was true that theoretically spot can be realised, practically it isn’t. For Paladin, about 10% is on contract as specified by SocGen, and the rest essentially has a guaranteed floor, a ceiling affected by spot and a realised price somewhere between the two.

Although there is demand for 62 reactors (on order/planned) and a further 159 proposed according to Areva, there is also a reality surrounding their construction. Considering each reactor costs about US$2bn to US$3bn (or ~US$2bn to US$2.5bn per GW), and like most things there is a shortage of quality metals to construct them, it could take up to 3 years to get the components. Fortunately, the 5 years of construction requires about 2 years of site preparation, so there should be some coincident timing. Consequently effective construction from approval could still take 6 or 7 years, (including ~6 months to commission [fortunately the commissioning period is short, not like a major base metal production plant]).

Most of the potential Australian uranium producers at the Paydirt conference appeared comfortable with only starting to producing uranium from ~2011 to 2013, because as shown in Figure 1, the main supply-demand gap appears to occur in 2014 after the Russian HEU agreement terminates, with current production building up from about 42,000tpa of yellowcake, and factoring in mine expansions and closures.

The Russian agreement with the US, relates to the sell down of their HEU stockpile of highly enriched (weapons grade) uranium that is diluted down (using depleted uranium) by 25 or 30 to 1 to reach acceptable enriched uranium reactor feed levels. The deal was based on an agreement in February 1993 in which Russia agreed to sell down 500tpa of HEU from dismantled nuclear warheads for 20 years to 2013 and effectively produce the equivalent of ~9,000tpa of yellowcake.

There also appeared to be an issue with enrichment capacity, which Areva was addressing, while Necsa was considering its own enrichment facilities if it takes the PBMR route.

Perhaps the reality check for the uranium industry will be when the profits are realised and the actually prices received either stated or determined from the revenues realised. However, that is currently sometime in the future. As part of that check would be the actual costs as almost everyone (like most mining producers) seems to be in the lowest quartile, which for uranium appears to be less than US$15/lb.

Disclosure and Disclaimer : This article has been written by Keith Goode, the Managing Director of Eagle Research Advisory Pty Ltd, (ERA, an independent research company) who is an Authorised Representative with Taylor Collison Ltd, and with his associates, may hold interests in some of the stocks mentioned in this article. The opinions expressed in this article should not be taken as investment advice, but are based on observations by the author. The author does not warrant the accuracy or completeness of any information and is not liable for any loss or damage suffered through any reliance on its contents.

Figure 1. Areva’s estimates of Reactor demand for tonnes of U3O8 to 2014 (Source: Areva, Mar 2007)GDNapr07

  • Written by: Keith Goode
  • Sunday, 01 April 2007