Twinkle, twinkle little Topaz


A nasty dispute between the Strategic Defense Initiative Organization (SDIO) and the space-science community was narrowly averted early this year when SDI officials backed down on a planned launch of a satellite with a Russian-built nuclear reactor on board. The scientific community, including the American Astronomical Society, complained that if SDIO tested the "Topaz 2" in 1995 as planned, the satellite would block the observations of a number of astrophysical satellites, destroying their ability to make scientific observations for years to come.

After angry words were exchanged, SDI officials agreed to delay the launch and to try to find a solution to the problem - probably by launching the satellite into a higher orbit than originally envisioned. But the dispute highlighted the complexities of using nuclear reactors in space and the controversy that has surrounded the practice.

The Soviet Union began putting small reactors into space in 1967. Early Soviet space reactors orbited for two to three months, and two Soviet Topaz 1 experiments, one for six and one 12 months, created radiation belts that disrupted the operation of scientific satellites. SDIO planned a two-to-three-year test of Topaz 2, and scientists complained that the effects of such an extended test would be even worse.

In the early years of the space age, compact and powerful nuclear reactors were expected to give mankind a universal source of energy. Both U.S. and Soviet designers thought they would have nuclear-powered launch vehicles within a decade. But those plans met with little success. Nuclear rocket engines proved to be nearly impossible to design and very dangerous to operate because of their extremely high working temperatures and the dispersion of highly radioactive materials along with engine exhaust. And satellites with nuclear-powered instrumentation have worked no better than those that rely on solar arrays to supply power for instruments.

In fact, nuclear power has only two potential advantages in space applications. The first involves traveling to the outer planets. Since the efficiency of solar arrays drops as the distance from the sun increases, nuclear reactors, which continue to produce the same amount of energy, offer some advantage. A second potential application is military—the use of nuclear power for satellites designed to survive a nuclear war in space. Nuclear reactors would be much less sensitive than solar arrays to the effects of nuclear explosions in space.

For some time, the U.S. government has supported its own "SP-100" program to develop a space reactor to serve just these two purposes. The SP-100 was to be a reactor with an output of 100 kilowatts. But the development of the device took much longer than expected, and the military began to lose faith in the program in the late 1980s. With the military withholding its portion of the SP-100 budget, the residual funding provided by the Energy Department and NASA is not enough to build a ground prototype, let alone launch it into space.

Meanwhile, by the late 1980s, the Soviet Union had developed a new type of small reactor—one pioneering the use of so-called "thermionic" conversion of heat into electricity. Thermionic conversion uses heat directly to emit electrons that create electric current. In contrast, thermoelectric reactors first absorb heat into a liquid metal coolant, and then transfer the heat to solid-state devices that produce electricity by the Seebek Effect. Since thermionic conversion is more direct, it promises a higher power-to-weight ratio and greater efficiency in converting heat to electricity. However, this may be true only at outputs well below the 100 kilowatt level that NASA has wanted for planetary applications. Furthermore, thermionic converters may have a much shorter lifetime than thermoelectric converters.

Consequently, NASA rejected the thermionic approach from the beginning. But the U.S. Air Force, when it assessed its power requirements in 1989, concluded that it did need a thermionic reactor. Meanwhile, the end of the Cold War made it seem that a Soviet-built reactor could fill the bill.

When Russian space hardware was put up for sale, space reactors were a primary target for the U.S. military, which claimed that a purchase would be, in part, a demonstration of international scientific cooperation. They focused their sights not on the flight-tested Topaz 1, but on a new unit designed by a competing contractor. Originally called the Enisey, the reactor was renamed Topaz 2 to inspire confidence in an imaginary lineage. The key argument in favor of the Topaz 2 was that it operated at lower temperatures than the SP-100 and would be more difficult to locate with infrared sensors.

The U.S. purchased two of these reactors in 1992. With their nuclear fuel replaced by electric heaters, the reactors were ground-tested at Albuquerque, New Mexico. This allowed the reactor system to be tested without disbursing radioactivity. After a successful evaluation, SDI officials decided to proceed with an orbital test.

But they hadn't counted on opposition. A major segment of the scientific community protested the plan, which would place the satellite in a low earth orbit. The NASA High Energy Astrophysics Mission Operations Working Group and the High Energy Astrophysics Division of the American Astronomical Society complained that radiation from the reactor would blind X- and gamma ray sensors on earth-orbiting astronomical satellites at distances as far away as hundreds or even thousands of miles.

Moreover, the test would result in secondary charged particles becoming trapped by the earth's magnetic field and forming artificial radiation belts. As scientific satellites crossed these belts, their operations would be disrupted. This problem would even render some of the space shuttle's orbits unusable. The damaging effects of these radiation belts had been demonstrated during earlier flights of Soviet nuclear-powered satellites, especially during the Topaz 1 tests in 1987-1988, which disrupted the U.S. Solar Max and Japanese Ginga satellites.

The immediate controversy was eventually resolved when SDIO agreed to reconsider the mission and find an acceptable alternative. In the meantime, SDIO is being disbanded and its research projects distributed among the military services. And cuts in the budget for space-based SDI programs may force further delays or a cancellation of the Topaz 2 test alternative. But the general problem of using nuclear power in space remains unresolved. Steve Aftergood of the Federation of American Scientists labeled the Topaz mission as the U.S. military's "probing balloon" to test the political acceptability of nuclear reactors in earth orbit.

Is a satellite with a nuclear reactor on board such a bad idea? Might it be possible to develop commercial space vehicles with reactor-powered instruments? It seems questionable. Even some SDI officials admit that the projected cost of reactor-powered satellites is higher than the cost of solar-powered satellites. Preliminary research would require at least $500 million more before reactors could be used in routine flight operations. Therefore, the commercial competitiveness of reactor-powered satellites is highly doubtful. Only military users would finance the development and use of such a power source.

SDI's mission was "to develop and test technologies for follow-up strategic defenses"; even if the program is transferred to the air force, it does not look as if this idea will be completely abandoned. And it is unlikely that the military would allow the choice of orbits for space-based military satellites to be decided by the scientific community. The Compton Gamma-Ray Observatory Users Committee has expressed concern that a successful test of Topaz could lead to the proliferation of nuclear-powered systems in high and low earth orbits.

The history of Soviet reactor-powered satellites shows that such proliferation may hurt more than space observatories. In January 1978 the Cosmos 954 reentered the atmosphere and fell in Canada. In 1983, as the orbit of another Soviet satellite, Cosmos 1402, decayed, the satellite spewed radioactive nuclear fuel and waste into the atmosphere. In September 1988, Cosmos 1900 lost control and narrowly escaped fiery destruction. Aside from these well known cases, two more satellites were reportedly destroyed in launch failures in 1969 and 1973.

Proponents of nuclear-powered satellites say they will be launched into "nuclear-safe orbits"—above 1,000 kilometers. But the safety of that altitude is relative. It is true that if would take several centuries before the satellite's orbit decays—ample time to complete technical projects. But the half-life of the plutonium 239 produced in the reactor is 24,000 years. This does not resolve the problem of highly radioactive waste falling on the earth; it merely leaves it to be resolved by future generations.

The problems of our descendants hundreds or thousands of years in the future probably do not concern many of us. But nuclear reactors in orbit present a more immediate danger. Low earth orbits, where potential military systems would function, is already densely populated by current satellites and debris from older spacecraft. A collision with one of the thousands of these pieces that circle the earth could lead to the destruction of reactors and the dispersion of their highly radioactive contents through space.

Many people are currently concerned about the possibility of the earth's colliding with an asteroid. Yet the probability of a space reactor colliding with a manmade object in near earth orbit is significantly higher than the probability of the earth colliding with an asteroid.

In 1988 the Federation of American Scientists and the Committee of Soviet Scientists for Peace Against Nuclear Threat urged their governments to ban nuclear power sources in earth orbit. The appeal received no official response, but there has been a de facto moratorium on launches of nuclear-powered satellites since then. In the United States, public opinion has been so supportive of the ban that environmental groups have tried to bar the flights of two nuclear-powered deep-space probes, the Galileo and the Ulysses, arguing that the threat of an accident during launch posed an unacceptable danger. The Soviet anti-nuclear movement has won an end to nuclear testing in Kazakhstan. Since Soviet satellites were launched from Kazakhstan, it is hardly feasible for Russia to resume such launches.

Incidents with nuclear space reactors have occurred every five years with such frightening consistency that after the incident in 1988 I joked that "we can relax until 1993" (although the editors of a Russian newspaper saw fit to remove the line from my article). My prediction has come true, but in an unexpected manner. This may be the year for decisions that will determine the future of nuclear reactors in space. If efforts are not made now to limit the space applications of nuclear power, we will run the risk of seeing a sky sparkling with Topazes.


July/August, 1993 THE BULLETIN of the Atomic Scientists