The Society for the Preservation of Canada’s Nuclear Heritage Inc. (SPCNHI) recently purchased two books about the heavy water plants in Cape Breton, Nova Scotia: “Hell and Heavy Water – A Glace Bay Heavy Water Plant Story” by Roland MacInnis (2017), and “Is Making Heavy Water Painful?” by Stan Davies (2023) about Port Hawkesbury. The authors, both former plant managers, tell a little-remembered part of Canadian nuclear history.
So what is heavy water? The nucleus of almost all hydrogen (H) atoms consists of only one proton, but about 1 in 6,400 also include a neutron. The heavy hydrogen, or deuterium (D), isotope has very useful properties for nuclear reactors – it is the best means to moderate or slow down neutrons flying out from split (fissioned) uranium nuclei, and slower neutrons are necessary to split additional uranium atoms. Heavy hydrogen also has the property or rarely capturing a second neutron. Without very pure heavy hydrogen (in the form of D2O or heavy water), CANDU reactors would not work.
To produce heavy water one has to extract heavy hydrogen from regular water, a challenging process exploiting small differences in the properties of light and heavy hydrogen. The first heavy water was produced in the 1930s at a hydroelectric plant in Vemork, Norway, making small quantities for universities and research laboratories. With the announcement of nuclear fission in January 1939, heavy water took on an increasingly important role, especially with the German and French nuclear programs, and later that of Canada. The Vemork plant was the target of daring commando raids during WWII, featuring in several books and movies.
During WWII, the British-European nuclear program moved to the Montreal Laboratories, which evolved into the Chalk River Nuclear Laboratories. Because of French and British expertise with heavy water, using 180 litres originally from Norway, the Montreal program naturally focused on heavy water. The USA Manhattan Project initially explored both heavy water and pure graphite as moderators. However, their success with graphite, which was easier to produce than heavy water, relegated heavy water to second choice as moderator. The US supported the heavy water work in Montreal, and provided a few tonnes of the precious liquid.
The post-war Canadian reactor program continued to work with heavy water, beginning with the ZEEP reactor (1945 Sep 5) and the much larger NRX (1947 July 22). For both reactors, the heavy water came from the USA; ironically, a few tonnes of D2O were produced annually in Trail B.C., from 1944 to 1956, in a plant similar to that in Norway, but the output was entirely purchased by the US atomic program. I recommend “An Early History of Heavy Water” by Chris Waltham (available online) describing both the Vemork and Trail plants.
The Canadian nuclear program needed hundreds of tonnes of heavy water for its large research reactors (NRX, NRU and WR-1) and its nascent CANDU power reactor program, yet did not produce any for its own use. The Trail plant produced 6 tons of D2O per year, at an astonishing $720,000 to $1,200,000 per ton (St John’s Evening Telegram, 1955 May 6, “In Near Future Atomic Power Stations Will Produce Nation’s Electricity)However, the Girdler sulphide (GS) heavy water extraction process had been developed at Savannah River, South Carolina, by Jerome Spevack in 1943; it was implemented in American industrial plants, enabling the expensive Trail facility to close in 1956. The US Atomic Energy Commission was able to supply heavy water for early Canadian reactor sales (Pakistan, Taiwan, India, Québec and Ontario) from their stockpile. However, the demand for heavy water was so great that Atomic Energy of Canada Limited even approached the USSR about purchasing some, and the experimental Gentilly-1 was shut down in 1972 to transfer its D2O to other reactors.
AECL wanted heavy water plants built in western Canada, where there was plentiful cheap fuel and lots of chemical engineering experience (i.e., refineries). But federal and provincial politics won the day, and economically-depressed Cape Breton was awarded two plants. Deuterium of Canada was located in Glace Bay – it was presided over by Spevack who, in 1963, persuaded the Nova Scotia and Federal governments to massively subsidize a plant. It used sea water for feedstock, leading to severe corrosion. Governments continued to pour money into the project but the only winner was Spevack – in 1968 he was dismissed with a pay out by the NS government, but the plant never produced any D2O on his watch. Eventually AECL was strong-armed into rebuilding it and, using fresh water as feedstock, it was finally commissioned in 1976.
The Port Hawkesbury (Cape Breton) heavy water plant was designed and built by Canadian General Electric; it was much more successful, using fresh water, although it still had considerable teething troubles. It came on line in 1971 and AECL was encouraged to take it over.
Ontario Hydro also produced heavy water, starting in June 1973, at its successful Bruce Heavy Water Plant (built by AECL). Initially it used boilers – fired by imported oil – with additional steam from the Douglas Point reactor until the Bruce A reactors started.
With all the heavy water production now in place, the inevitable happened; demand fell along with reactor sales, and all the plants were shut and demolished. Today Canada does not produce any heavy water, relying on foreign suppliers, accumulated stocks, and purification facilities to upgrade used D2O. Typical 2023 prices for heavy water are $US 110,000 per tonne.
If you would like to borrow our new heavy water books, or peruse our large collection of books and reports on Canada’s nuclear history, end us an email at info@nuclearheritage.com.
