BREEDER SAFETY

In a world where "news" is rumor leaking out from a tight straitjacket of secrecy, such statements cannot be conclusively verified; but we find the story perfectly credible. That the wastes should have undergone a nuclear explosion is, of course, patent nonsense. But that they were buried shallow underground, overheated, and turned moisture into steam which then explosively erupted and spread radioactivity is not only physically possible; it also fits the general pattern of Soviet safety measures.

The approach to safety is well illustrated by the breeder reactor, which along with coal is on of the major solutions to the energy problem for the rest of this century.

The plutonium "issue" is a red herring, among other reasons because 1) the amount of plutonium in the power industry will, far into the future, be small compared with that in the weapons industry, where no problems have arisen in over 30 years; 2) the amount of plutonium bred in a breeder can be precisely controlled, so that the amount of surplus plutonium outside a reactor (either being bred or used as fuel) at any time can be kept to a minimum close to zero; 3) if the scaremongers were really so concerned about plutonium, they could advocate the thorium cycle, which breeds non-explosive uranium 233 rather than plutonium ever heard them?

The more important problem is in the heat exchangers. In Britain and Germany, the nuclear parts of fast liquid-metal breeders have been ready to go for some time, but there have been problems with the sodium/water heat exchangers, a problem successfully solved in France. (The US has an experimental breeder in Idaho Falls; but as a public relations stunt, Carter has severely cut the budget of the US breeder program.)

Why is the breeder "fast"? It isn't; what is fast are its neutrons. In a light-water reactor, the neutrons emanating from the fissile material (U 235) are slowed down by a moderator, for only slow neutrons will cause further U 235 nuclei to split. In a breeder, they must keep their speed, for only fast neutrons will turn fertile material (U 238) into Missile plutonium. That excludes using water as a cooling and convection fluid, for its hydrogen nuclei would slow down ("moderate") the neutrons.

The fluid, then, must be an element whose nuclei do not moderate the neutrons, and which has acceptable thermal properties, including a high boiling point (obviating the need to pressurize it). The element that best suits these (and other) specifications is liquid sodium. That is why the breeder is "fast" and "liquid metal."

But sodium reacts violently with water, and the heat exchanger transmitting the heat of the breeder-heated sodium to the turbine-driving steam must not allow the two to come into physical contact.

The two countries that now have a breeder actually on line, France and the USSR, have taken very different approaches to the problem. The highly successful French design is well described in the March issue of the Scientific American. It reduces the probability of a fault to a low value and contains the reaction if a fault develops nevertheless. The French 250 MW Phenix breeder has been on line since July 1974, and has been producing electric power commercially with a reliability as good or better than fossil-fired plants, and with a higher efficiency (43%). The 1,250 MW Superphenix is planned to go on line in the mid-1980's.

In contrast, in the Soviet BR5 breeder at Shevchenko, the only thing that stood between the sodium and the steam was some tubing welded by "Soviet technology the world's most progressive." Go on line now, fret later, is the Soviet safety policy, and the world's most progressive welds went bust almost as soon as the contraption went into operation in 1973; the crippled reactor is now on line with about one third of its original 350 MW capacity.

As usual, there are no details, and one is left to one's own imagination to picture how the welders were banished to Siberia, the posthumous decorations handed out, and a fresh ox-cart load of plutonium hauled in.