Why Was Diffusion Selected for Eurodif?

[Delta Force]

The Eurodif gaseous diffusion plant required so much electricity that three of the four 955 megawatt nuclear reactors at Tricastin were allocated to meet to its power demands. In fact the Eurodif plant was such a large industrial consumer that the reactors were were built close by to minimize transmission losses, and as of 2012 it was consuming the equivalent of 4.5% of France's total electrical demand. After Eurodif switched to using gas centrifuge technology at least two of the reactors were freed up to provide power to other customers.

Nuclear power plants were less expensive in the 1970s and electricity demand was rapidly growing at that time, so building a massive nuclear power station to meet the needs of a single industrial consumer might not have seemed as radical at the time. For example, the Kingston coal fired power station was built to help meet the needs of Oak Ridge, and the Midland cogeneration plant in the United States was originally going to be a dual unit nuclear power station (it was completed as a natural gas plant) to provide heat and power to nearby chemical factories. Still, 3.6 gigawatts is a massive amount of power. Gas centrifuge technology was also undergoing rapid development for Pakistan's nuclear weapons program, and many of the components were imported from European suppliers.

Did the Eurodif project ever consider alternative enrichment technologies such as gas centrifuges or chemical enrichment (another French enrichment technology)? Furthermore, since the 1970s was a time of growth in nuclear power were there concerns that newer enrichment technologies might overtake Eurodif?

 

[Astronuc]

This IAEA document describes the state of the art in enrichment by 1977.
https://www.iaea.org/sites/default/files/publications/magazines/bulletin/bull19-1/19104884052.pdf

Experiments on the use of centrifuges for the separation of mixtures of gases of different molecular weight were carried out at the end of the 19th century. An experimental centrifuge plant for separation of uranium isotopes in UF6 was operated in the 1940's. However, it was abandoned in favour of gaseous diffusion mainly because of problems with materials and bearings which were unable to withstand the very high rotational speed and centrifugal forces required.

The development of stronger and lighter materials, of new bearings," and of improved designs of centrifuges has brought the centrifuge isotope separation back into use in Europe and the USA. Experimental work is also being done in Australia and Japan. The attraction of the centrifuge method is that, at its present stage of development, it permits separation per stage about two orders of magnitude larger than with other methods in use, and that the energy consumption is only about 10% of that of the gaseous diffusion process. A disadvantage is that much smaller quantities can be processed in each centrifuge than in a gaseous diffusion stage. Consequently very many centrifuges must be operated in parallel and the total number of centrifuges therefore must be much larger than the number of separation stages in the gaseous diffusion process in order to produce the same output.

Centrifuge technology was not mature at the time. High strength materials and high speed bearings were necessary.

On page 48 of the IAEA document:

Centrifuge process.

This process is new on the market. It was developed concurrently in the Federal Republic of Germany, the Netherlands and the UK. By the Almelo Agreements
of 1970, the three Governments established technological and industrial co-operation on enrichment. This co-operation is executed by two companies (URENCO Ltd. and CENTEC GmbH) and two enterprises (URENCO UK and URENCO NEDERLAND). The two enterprises are responsible for design, construction, ownership and operation of centrifuge enrichment plants at Capenhurst in the UK and at Almelo in the Netherlands, respectively. A third enterprise in FGR is possible. URENCO provides a central marketing service while CENTEC co-ordinates the joint tripartite research and development programme. All companies have shareholders from all three participating countries.

Three pilot plants with a total capacity of 150 000 SWU per year are in operation, a Dutch plant and a German one at Almelo, Netherlands, and a British plant at Capenhurst, UK. Experience from these plants is being used for construction and operation of two production plants, one at Almelo and one at Capenhurst, which started operation in 1976 and will each be operating at full capacity of 200 000 SWU per year in 1978. URENCO will utilize the flexibility of the centrifuge process to progressively install further capacity to reach a total of 2 million SWU per year by 1982.

 

[Delta Force]

This IAEA document describes the state of the art in enrichment by 1977.
https://www.iaea.org/sites/default/files/publications/magazines/bulletin/bull19-1/19104884052.pdf

AREVA says Georges Besse I (the gaseous diffusion plant, Georges Besse II is the gas centrifuge replacement) had a total capacity of 10.8 MSWU per year, but only produced 150 MSWU throughout its entire service life. That gives an average output of 4.17 to 4.55 MSWU per year (diffusion ran from 1976 or 1979 to 2012). Was Eurodif massively overbuilt in anticipation of future nuclear fuel demands that never arouse or did Urenco and other firms using gas centrifuge technology eat into its market share in the limited nuclear fuel market that ultimately arouse?

Also, were the three reactors at Tricastin allocated to Eurodif for its total capacity (in practice resulting in a surplus for the grid due to reduced operations) or just what was necessary to meet the reduced operations that ultimately transpired?

Centrifuge technology was not mature at the time. High strength materials and high speed bearings were necessary.

So at that point in time it was deemed less expensive to use ten times the electrical capacity and go with gaseous diffusion instead of going with centrifuges? Do you know how much of that is due to power plant capacity (especially nuclear) being so inexpensive at the time versus how much of that was due to centrifuges being unreliable?

 

[Astronuc]

AREVA says Georges Besse I (the gaseous diffusion plant, Georges Besse II is the gas centrifuge replacement) had a total capacity of 10.8 MSWU per year, but only produced 150 MSWU throughout its entire service life. That gives an average output of 4.17 to 4.55 MSWU per year (diffusion ran from 1976 or 1979 to 2012). Was Eurodif massively overbuilt in anticipation of future nuclear fuel demands that never arouse or did Urenco and other firms using gas centrifuge technology eat into its market share in the limited nuclear fuel market that ultimately arouse?

From the AREVA site referenced: "Located on the Tricastin site, the Georges Besse II plant whose construction was started 3 years ago will eventually be made up of 2 units of enrichment. Thanks to its modular nature it will reach a total production of 7.5 million SWU per year in 2016, 2 years earlier than originally planned.The Georges Besse II plant produced its first SWU in December 2009."

Gaseous diffusion technology apparently lead centrifuge technology, so GD was selected for industrial scale in the 1970s. I don't know if the George Besse plant was overbuilt, but I suspect it was built with the expectation that Europe would have more nuclear reactors, and such large capital projects are driven by economies of scale. The GB I was in operation when TMI-2 had it's accident, and consequently, plans for additional NPPs were suspended or cancelled. Given that GB I was up and running in 1977, they probably had 30 years of operation in mind for recovery of investment.

Had centrifuge technology been perfected by the early 1970s, I would imagine that Eurodif would have gone that route. On the other hand, Urenco was able to introduce centrifuge technology in 1982, and the Capenhurst GD plant was shutdown after operating since the early 1950s.
http://www.wmsym.org/archives/1992/v1/150.pdf

According to a Wikipedia article, gas centrifuge technology was applied on the industrial scale fairly recently in the US. Prior to that, GD was the predominant technology at 3 sites: Paducah, Portsmouth and K-25.
https://en.wikipedia.org/wiki/Portsmouth_Gaseous_Diffusion_Plant

The economy of NPPs follows a similar trend. Basically, NPPs have been initially designed for 40 years with an expectation that they would be paid off at 30 years. Now, various plants have life extensions of 20 years, to a 60-year lifetime, and there is some effort to extend life to 80 years. Of course, some plants proved very expensive with substantial cost over-runs following the Browns Ferry fire and TMI-2 accident, which necessitated major design changes to ensure safety.