It appears from this paper, India loaded a PHWR with thorium in 1992: Surendra Mishra, R.S. Modak, S. Ganesan,
Optimization of Thorium loading in fresh core of Indian PHWR by evolutionary algorithms, Annals of Nuclear Energy,
Volume 36, Issue 7, 2009, Pages 948-955.
From the text:
Since thorium is a much stronger absorber than depleted Uranium, it causes stronger flux depressions in its vicinity. This affects the reactivity worth of the two shutdown systems. On the whole, the number and locations of Th bundles have to be chosen so that the following conditions are fulfilled:
(1)
Keff is maximum possible so that reactor can operate without fuelling for longer time leading to better fuel economy.
(2)
The operating limit on channel outlet temperature (299 °C) is obeyed.
(3)
Maximum bundle power is less than 440 KW.
(4)
The reactivity worth of each shutdown device is not less than 30 mk.
The problem of obtaining an optimum distribution of Thorium bundles which satisfies above criteria was solved successfully about two decades ago (Balakrishnan and Kakodkar, 1994) and a suitable distribution consisting of 35 Th bundles was arrived at. This pattern was subsequently loaded in the Indian PHWR at Kakrapar (KAPS-1), which went critical in September 1992. A gradient based method was used to find the solution.
My overall impression is that the Indians intend to use plutonium rather than HALEU to provide the initial reactivity. I think this a superior approach since HALEU relies on enrichment, with which I believe we should do away, since it adds expense.
I discussed my views this point, peripherally, on another website, at which I was a "guest author in 2013, with some reference to Indian CANDUs:"
On Plutonium, Nuclear War, and Nuclear Peace
Note that the 1992 experiment was conducted in a relatively small PHWR, not exactly of the CANDU design. It used natural uranium, like a CANDU, but as the text excerpted hints, reaching full power took some time because of the higher neutron capture cross section of
232Th compared to
238U. The longer half-life of the intermediate nuclide in the thorium case,
233Pa (t
1/2 = 26.975 days) as compared to the intermediate in the
238U case,
239Np (t
1/2 = 2.356 days) means that the fissionable material takes longer to accumulate in the former case, and thus longer to reach full power.
It seems to me, without any sophisticated analysis, just off the top of my head, that a ternary Th, U, Pu fuel would not suffer from this lag. My son keeps promising to get me access to publicly available neutronics programs, but the little brat hasn't done so yet. He doesn't work with that software himself, as he's in nuclear materials, not fuels, but his girlfriend and one of his best friends do. There are a few zillion examples I'd love to run, but my computational power, and time, are limited.
= new reply since forum marked as read