Who is afraid of breeders?

We present here our comments on the article by Tongia and Arunachalam1.

Who is afraid of breeders? Many are. The main basis for the fear is the concern for nuclear proliferation as the fast breeder reactor (FBR) breeds or produces more fissile material than it consumes. The fear has manifested itself in many forms.

One of the earlier manifestations was the argument against the closed U–Pu cycle initiated during the International Fuel Cycle Evaluation (INFCE)2. Fortunately and correctly, the conclusion from the INFCE studies was that ‘No fuel cycle is free from proliferation’. Nevertheless, we witnessed the abandoning of the FBR programme in the UK, Germany and USA; the joint European FBR programme also has been disbanded all due to the fear. Then came the campaign that as there are abundant resources of energy (coal, oil, gas, uranium and the highly enriched uranium and plutonium from the dismantled weapons), there is no need to breed, we should only burn fissile material. So France is converting its breeders to burners, with inert matrix fuels, etc.3 The Russian Federation, Japan and India are the only three countries that have announced plans to continue with the FBR on a commercial scale. (While China is now building a 65 MW test fast reactor, and South Korea a 50 MW test fast reactor, we started construction of our 40 MW Fast Breeder Test Reactor [FBTR] way back in the seventies.) Now come Tongia and Arunachalam with the advice to India that fast breeders do not breed fast enough and therefore ‘India should consider entering into long-term agreements with other countries, with appropriate policy innovations for importing uranium’. Alas they have not told us which countries and what are the ‘appropriate policy innovations’. It is interesting that in their paper under Methodology, in the sub-section ‘Plutonium from PHWRs’ (p. 553), Tongia and Arunachalam state, ‘The availability of uranium from other countries is not included in these calculations as there are restrictions imposed by the Nuclear Suppliers Group on the supply of nuclear materials to India’. So their advice is very

997.jpg (9344 bytes)


Figure 1.  Installed electrical capacity in India and possible contributions from oxide FBRs (0.6 PLF, 2% cycle losses) and metallic FBRs (0.6 PLF, 1% cycle losses).

Figure 2.  The learning curve for technology development.

clear. The appropriate policy innovations should be those that satisfy the Nuclear Suppliers Group to remove the restrictions!!

Fast breeders do not breed fast

The main conclusion in their paper is that it appears difficult to have FBRs with short doubling times and hence the contribution of FBRs to the country’s total electric capacity requirement could remain low till the latter half of the 21st century (figures 5 and 6 of their paper).

It is a common misconception that the adjective fast in fast breeder qualifies the word ‘breeder’. It does not and indeed fast breeders do not breed fast. Walter Marshall4, in the context of ~ 30 years doubling time for the oxide fuel, commented in 1980, ‘In short, fast breeders do not breed fast, they simply use fast neutrons and breed rather slowly’. But as the fissile atom density in the fuel increases, i.e. as we go from oxide   carbide   nitride   metal, the doubling time decreases and fissile material growth rate increases (as indeed are the results of Tongia and Arunachalam in their tables 4 and 7 respectively). They have also demonstrated the effect of better breeding of advanced fuels (metal/carbide) in their figures 5 and 6. It is surprising that in their Figure 4 a they plot the possible nuclear capacity for oxide FBRs only. In Figure 1 we depict their Figure 4 a including the possible FBR nuclear capacity with metallic fuel, using data given by them. Even with their questionable assumptions; the growth rate is good!

Dangers in projecting from the past and present to the future

As scientists and engineers, all of us know the dangers in extrapolating to the future using data from the past/present. The slope in the future of the learning curve of any new technology (Figure 2) is difficult to predict and the pitfalls in making long-range predictions are obvious. In the analysis by Tongia and Arunachalam plant load factor (PLF) is the most important parameter and they vary it between 40 and 70%. True, these were the PLFs for Indian PHWRs before 1995. But TAPS-I had PLF of 84%, MAPS-II of 78% and Narora of 90% in 1997–98 (ref. 5). The average PLF for all the plants in the last three years is greater than 70%. Indeed with these PLFs even with the many other erroneous assumptions in their paper, nuclear growth through FBRs becomes feasible particularly with advanced fuels. It is wrong to project the low PLF of the earliest years as valid for the growth of FBRs for over a century in the future.

Resource utilization

FBRs have many advantages. In the context of the paper under discussion, two are relevant and important: growth capability and resource utilization capability. The authors have failed to note that the growth capability of FBRs is secondary to their resource utilization capability. By the use of FBRs the utilization of uranium can reach 60–80% as compared to less than 1% with LWRs and PHWRs on once through cycle or a few percent with Pu recycle6 (Figure 3). As noted in table 1 of the paper by Tongia and Arunachalam, this translates to atleast 50 times more energy

998.jpg (10796 bytes)

Figure 3.  Utilization of uranium as a function of conversion/breeding ratio.

extracted from indigeneous uranium or 50 times less uranium to be imported for the same amount of electricity produced. In Figure 4, we give the potential of different energy resources in India. For energy independence, the long range planning should base a major share of electricity generation of a country on indigenous resources. The importance of FBRs, in this context, for India is obvious, given the kind of growth in energy demand expected in the coming decades and later into the second half of the next century and even for a lower growth rate than that assumed by Tongia and Arunachalam.

The question is not where will the uranium come from but where is the uranium?

While the paper has recommended the construction of a large number of thermal reactors using imported uranium to rapidly increase the nuclear power base, it has not studied the implications of the suggested strategy and the kind of electrical capacity growth that could be achieved (as in their figures 5 and 6). Without such an analysis the recommendation of the paper cannot be accepted. In their figure 4 it is seen that by 2055 a target electric capacity of 600 GWe is projected, and if as advocated in the paper, a substantial fraction (say 50%) of this is to be from thermal nuclear power plants, then it implies that about 300 GWe should be fed by imported uranium. On once through fuel cycle basis this requires mined uranium of 20 times the U resource available in the country, i.e. 1.2 million tonnes of natural uranium, which is more than the reasonably assured uranium reserves in USA and Canada together. Not only is such import unfeasible in the context of world uranium availability but such an import requirement will grossly compromise the energy independence and energy security of India. In addition, the strategy for reprocessing or other disposition of the spent imported fuel requires considerable study and is not touched upon in the paper. It is also to be noted that the use of Pu–Th cycle in thermal reactors cannot increase the installed thermal nuclear power capacity but can at most maintain it at the same level.

A recent OECD study7 has examined three scenarios for the change in the present nuclear share in the world energy production in the next hundred years (Table 1 and Figure 5). The most optimistic scenario (I) projects a total nuclear installed capacity of 1120 GWe in 2050 for the whole world, which means a production of about 1 TWy per year. We cite these only to show that a total nuclear generation of 1 TWy per year can be sustained by the open LWR cycle for only 44 years by the world’s uranium resources8 (Figure 6). In the absence of any other new technology, FBRs will ultimately become relevant and necessary for the whole world.

FBR initiation does not depend on doubling time

Starting from the fact of long doubling times of present day FBRs, Tongia and Arunachalam conclude that the Indian FBR programme will take a very long time to make a significant contribution to the growing national electricity base. This conclusion come from incorrect or unjustified assumptions as explained below. The natural uranium available in India can be used to set up about 10–15 GWe of PHWRs. The Pu available from the spent fuel of this initial PHWR base is enough to set up over 25 GWe of FBR capacity, i.e. atleast 50 FBRs of 500 MWe capacity each. The growth rate upto this initial FBR capacity is governed by the rate of setting up of PHWR plants and the rate at which the PHWR spent fuel is reprocessed. In this period the fissile growth from the FBR breeding process is only a bonus and certainly not a limiting factor, i.e. in this phase it is the Resource Utilization Capability of the FBRs which is important and not the growth capability.

999.jpg (19591 bytes)


Figure 4.  Comparison of different energy resources for electricity generation in India.

Figure 5.  World nuclear electricity generation projections.

Table 1.  Three variants of world nuclear power capacity (GWe) up to 2050

The need for short doubling time arises only after a substantial base of FBRs (25 GWe) is set up and hence only compound system doubling time (CSDT) applies and the simple doubling times (SDT) used in the paper are not applicable.

We are working on advanced fuels for the future

Once the initial FBR capacity based on PHWR Pu has been successfully established, it will be necessary to have FBRs of short doubling time (12–15 years) to be able to further grow the FBR capacity in tune with the growth of the national electricity base. Many studies exist defining the technological developments needed in order to enable FBRs to have short doubling times. In particular, Tongia and Arunachalam have referred to studies by Lee who has identified several required developments to current FBR technology in order to have adequately short doubling times in the Indian context. In fact, we are very much aware of the need to develop advanced FBR fuel (advanced oxide, nitride, carbide or metal alloy) in order to have FBRs with good growth capability. R&D for advanced FBR fuels is one of the important objectives of the Indira Gandhi Centre for Atomic Research (IGCAR). Starting from the base provided by the mixed carbide fuel experience in FBTR, studies on advanced fuels including their fuel cycle are a part of the IGCAR R&D programme. The extreme sensitivity of the doubling time to factors such as fuel material, fuel design (e.g. pin diameter), plant load factor, out of pile cycle time, discharge burnup, cycle fissile losses and breeding gain is also well known. Studies2 have shown that with a large pin diameter optimized for breeding, even the oxide fuel can give doubling times as low as 15 years and this is still lower for carbide/nitride/metal.

We note that Tongia and Arunachalam have made no attempt to calculate the values of FBOC and FG (in their eq. (1)) and take them from studies of others. On the other hand, pessimistic assumptions are made for EF, FL, PLF which are assumed to be applicable for over a hundred years in the future. In fact discharging fuel to out-of-pile storage and reprocessing of short cooled fuel are methods for reducing EF for future reactors. Reduction of FL is also very important for future FBRs to have short doubling times and to reduce the actinide waste storage problems.

1000.jpg (5847 bytes)

Figure 6 : world fixed energy resources (TWy)

Further, the plant load factor cannot be taken at low values for growth studies (p. 552). When PLF is very low the reactor system cannot be sustained on economic grounds itself and any growth capability is irrelevant. Hence it is necessary to take the minimum PLF at the level needed for economic viability as the input for growth capability studies.


In conclusion the fast reactor system remains the only system for a complete

and meaningful utilization of India’s nuclear resources. Without this
system it is not possible to have a
sustainable nuclear power develop-
ment in the country. As coal resources deplete and hydro resources saturate, the nuclear power fraction has to increase. The FBR option remains the only means of meeting the electrical energy requirements in the indust-
rially developed India of the 21st century. Import of natural uranium along with light water reactors, can at best
be a short-term measure of meeting
immediate electrical energy demand
but cannot be recommended on a large scale or as an option for sustainable development.

  1. Tongia, R. and Arunachalam, V. S., Curr. Sci., 1998, 75, 549–558.
  2. Fast Breeders, Report of INFCE Working Group 5, IAEA, Vienna, 1980.
  1. International Symposium on Nuclear Fuel Cycle and Reactor Strategies, IAEA,
    Vienna, 2–6 June 1997.
  2. Marshall, W., Atom, 1980, 287, 222–
  3. Nu Power, 1998, 12, 2–3.
  4. Status of Liquid Metal Cooled Fast Breeder Reactors, Technical Reports
    Series No. 246, IAEA, Vienna, 1985,
    p. 3.
  5. Nuclear Power and Climate Change, Report of Nuclear Energy Agency, OECD, 1997.
  6. Walter, C. E., International Symposium on Nuclear Fuel Cycle and Reactor Strategies, IAEA, Vienna, 2–6 June 1997.


ACKNOWLEDGEMENTS.  We thank Mr S. B. Bhoje, Mr S. C. Chetal and Mr S. Govindarajan for providing various inputs needed for the analysis reported above. Thanks are also due to Mr S. E. Kannan and Dr S. Venkadesan for help.



Indira Gandhi Centre for

Atomic Research,

Kalpakkam 603 102, India