Update: This is one of the most popular posts on this blog and so merits an update. Since 2008 Canada’s nuclear situation has radically changed. There are now no prospects for more domestic CANDUs and any more CANDU exports are also very doubtful. The Quebec reactor, Gentilly II, has been shut down and in addition to the two reactors already shutdown, the remaining six at Pickering will be taken out of service by about 2020. Therefore, lots of heavy water will be available in future but likely no more new CANDUs to use it.
Heavy water (deuterium oxide) is fundamental to CANDU reactors. For example, a CANDU 6 reactor requires 265 Mg (metric tons) of heavy water for its moderator and 192 Mg for its heat transport system (coolant) making a total of 460 Mg per reactor. In comparison a Darlington reactor needs 592Mg. The Advanced CANDU Reactor (ACR-1000) will require 250 Mg all for its moderator but none for its coolant which will be light water.
Approximately 3Mg of heavy water are needed to make up the annual losses in operating CANDU reactors. For example, Ontario Power Generation (OPG) is committed to providing Bruce Power with 18 Mg to make up losses in the reactors it leases. The 18 Mg was based on the 6 reactors operating at the time of the agreement.
The last heavy water production plant in Canada, at the then Ontario Hydro Bruce site, closed in 1997 and has since been decommissioned. A 1 Mg per year prototype plant in Hamilton Ontario operated for about two years ending in 2002. Since then there has been no heavy water production plants in Canada nor have plans for new plants been announced.
Heavy water may be available from sources outside Canada principally India which produces about 600 Mg per year mainly for its own heavy water reactors.
The financial statements in Atomic Energy of Canada Ltd (AECL) Annual Reports for the past few years list an asset of $300M as “heavy water inventory”, mostly in storage at a site in LaPrade, Quebec adjacent to the Gentilly reactors. If the price of heavy water is assumed to be $600/kg, a reasonable ballpark number, then the asset would consist of 500 Mg and $300/kg would give 1,000 Mg. However, these simple minded calculations are unlikely to be correct because of the eccentricities of government book keeping in terms of asset valuation as the next paragraph shows.
Notes to the AECL financial statements indicate that the $300M includes the 1,003 Mg belonging to AECL on loan to the Sudbury Neutrino Observatory (SNO), a physics experiment. Therefore, the heavy water price used to value the inventory must be $300/kg or less.
Paraphrasing a recent OPG document, OPG owns 14,440 Mg of heavy water, of which 13,440 Mg is radioactive, and 1,000 Mg is non-radioactive. Most of the radioactive heavy water is in OPG (6,300 Mg) and Bruce Power (6,000) reactors; the other 1,140 Mg of radioactive heavy water are available in the closed down Pickering 2 and 3 reactors. With a provision of 500 Mg for make-up to operating reactors OPG also sells modest quantities from its stocks and leases some heavy water to AECL and other utilities.
At this time the maximum potential for building ACR-1000 reactors in Canada would appear to be two in Ontario, two in Alberta and one in New Brunswick for a total of five. It would appear that OPG’s existing stock of heavy water could cover the 1,250 Mg required, provided other arrangements for sale or lease don’t complicate the matter. However, it is doubtful that sufficient heavy water would be available to supply five Enhanced CANDU 6 reactors requiring 2,300 Mg.
The foregoing contains too much guess work and speculation to form a prudent basis on which to order heavy water reactors.
The proponents of CANDU reactors need to clarify the issue of heavy water supply by publicly quantifying the inventories they have on hand.