Washington Post
December 27, 1999
Pg. 25

Why The Test Ban Is Safe For Us

Don't let a vital agreement be killed by quibbles.

By Richard L. Garwin

In his op-ed article of Nov. 23, former secretary of state Henry Kissinger expresses concern about verification of the Comprehensive Test Ban Treaty (CTBT), the ability to maintain the U.S. stockpile of nuclear warheads and the lack of "meaningful sanctions" for violating the CTBT. I believe that an extensive set of Senate hearings--not the rush to judgment that we got--would have resolved these issues in the minds of many.

In 1950 I began work at the Los Alamos National Laboratory on fission weapons, innovations in nuclear explosion testing and early thermonuclear weapons. I served the Kennedy, Johnson and Nixon administrations on the President's Science Advisory Committee and in other roles dealing with national security technology and arms control. When Kissinger was national security adviser to President Nixon, I met with him frequently as a member of his small, informal technical advisory panel. In recent years, I have studied CTBT issues for the Department of Energy (DOE), especially our ability to maintain a stockpile of nuclear weapons of existing type that is safe and reliable without nuclear explosion testing.

Regarding verification: The CTBT's international monitoring system will detect all global events that release energy equivalent to the detonation of 1,000 tons or more of high explosive, and will locate these events within a radius of 18 kilometers.

Sensors of low-frequency sound in the atmosphere and oceans and detectors of airborne radioactivity will complete the system. Existing seismic detection capabilities that will supplement the CTBT's are in many cases far more sensitive than the minimum planned for the CTBT. For instance, existing seismo-meters in Pakistan would have detected additional tests of 10 tons' explosive yield at the Indian test range in May 1998; and seismometers in the Nordic countries often could detect an explosion of one ton at the Russian test site at Novaya Zemlya.

A country that cares about its role in the community of nations would not lightly violate a CTBT in the hope that it could evade detection. Efforts to evade detection by exploding the nuclear weapon in a large underground cavity are particularly chancy for a nation without testing experience.

The ability to keep U.S. nuclear weapons (now numbering some 10,000 weapons of about 10 types) safe and reliable under a CTBT depends on an improved surveillance program for the health of the weapons in the stockpile, as has long been practiced. Each weapon type in the stockpile has had its production verification test and would not normally be fired again no matter how long it remained in the stockpile. But every year, 11 units of each type are randomly withdrawn from the stockpile and inspected--one of them thoroughly disassembled to detect corrosion or other problems of aging.

Now as in the past, incipient deterioration would result in replacement of the offending item throughout the affected stockpile: gaskets, lubricants, etc. For the few components that can be operated only during a nuclear explosion, periodic remanufacture of the plutonium primary nuclear explosive in the thermonuclear weapon, or the uranium and lithium-deuterium secondary, would maintain the reliability of the stockpile without degradation, if the remanufactured nuclear components fall within the range of variability of the original manufacturing run. The CTBT question is not whether the stockpile is perfectly reliable but whether it is as reliable as it was in the days of nuclear testing.

Under the $45 billion Science-Based Stockpile Stewardship program, enormous strides in computing, in measurement of purely high-explosive tests and in laboratory experiments dealing with the transport of radiation already serve to increase our understanding of the stockpile beyond what it was in the old days. As the program matures, confidence in reliability should increase, but only if the DOE has the requisite manufacturing capability, which is now beginning to function. Adequacy of the manufacturing complex is not a CTBT question, but it deserves more attention by DOE.

The "traditional" means of maintaining a reliable stockpile by the substitution of new types of nuclear weapons (each with a single production verification test) in no way improves the reliability of the stockpile. Instead, the historical concentration of serious stockpile problems within the first four years of a weapon's presence in the stockpile implies that constant introduction of new designs increases the likelihood of stockpile problems.

As for Kissinger's plaint about the "lack of meaningful sanctions" in the CTBT: Sanctions are rare, almost to the point of nonexistence, in such treaties, including those for which he was responsible. Violation of one treaty or norm can bring retaliation in other domains.

The search for agreement on "precise and tough sanctions" is a prescription for indefinite delay that will imperil not only the CTBT but the Non-Proliferation Treaty and the web of constraints on the acquisition of weapons of mass destruction.

The writer is Philip D. Reed senior fellow for science and technology at the Council on Foreign Relations.

Editor's Note: The op-ed referred to appeared in the Current News Early Bird, November 22, 1999.