(2 of 3)
As long as a coolant (water in most reactors) keeps flowing around the reactor core, it carries heat away, and the temperature stays under control. If the coolant is lost, the core begins to overheat, like a car with a broken radiator. The chain reaction promptly ceases because rising temperatures cause the fuel to expand, which increases the distances between individual atoms and makes it less likely that the neutrons emitted by one will hit the nucleus of another. But the spontaneous radioactive decay of nuclei goes on. The uncooled reactor core could eventually get hot enough to melt through its casing and the surrounding building, causing fires that loft radioactive material into the atmosphere. Under the worst circumstances, the core melts through the earth and in a "China Syndrome" reaches the underground water table and triggers the further release of radioactive particles. In an effort to minimize the chances for such disasters, Sweden is developing the PIUS (for Process Inherent Ultimately Safe) reactor, which is immersed in a giant pool of water. If the primary cooling system on PIUS fails, pool water floods and cools the core. A reactor being developed by General Electric, Rockwell International and Argonne National Laboratory is cooled directly by submersion in a pool, except that the liquid is molten sodium, which can absorb far more heat than can water before boiling away. Still, should some accident -- an earthquake, for example -- empty the pool, these reactors could conceivably melt.
The MHTGR, in contrast, has no safety cooling system at all; the helium gas flowing through its core merely carries away heat to power electric generators. The reactor itself can never get hot enough to melt down. In the MHTGR, bits of uranium fuel are encapsulated in tiny grains made of carbon and silicon compounds. The fuel particles, which are embedded in racquetball-size "pebbles" of graphite, will remain intact up to 3600 degreesF. But the configuration of the core and the reactor's size (it generates only 80 megawatts of power, compared with 1,000 megawatts for large conventional reactors) ensure that temperatures never rise above 2900 degreesF. The MHTGR has another advantage, says Lidsky: its principal components could be mass- produced. Utilities could combine the outputs of several separate 80- megawatt modules to make one large plant.
With a concentrated effort, Lidsky argues, the MHTGR could be on-line in the U.S. by 1996. He has received some support from an unexpected source: the Union of Concerned Scientists, which is generally opposed to nuclear energy. Robert Pollard, a nuclear engineer and U.C.S. spokesman who has urged the shutdown of all U.S. nuclear plants, says that the MHTGR is a "much better idea than current reactors. It's basically a much slower-acting machine."
The U.S has already had a bit of experience with gas reactors. Philadelphia Electric Co. successfully tested a 40-megawatt experimental version from 1967 to 1974. However, the Fort St. Vrain plant, 35 miles north of Denver, has had one breakdown after another during the decade since it began operation. But Lidsky points out that the plant is so big -- 330 megawatts -- that it needs as complex a cooling system as conventional plants. The reactor's large size, he says, has caused most of the trouble.