Despite President Carter's professed desire for a nuclear test ban treaty, his own Department of Energy has warned Congress that a comprehensive ban longer than three years could endanger weapon reliability.
In testimony released yesterday, the agency presented Congress with a list of instances in which it said weapons in the U.S. nuclear stockpile would have failed to function property if they had lain untested for a number of years.
DOE which usually describes the stockpiled weapons as 98 percent reliable, outlined the series of problems in testimony designed to show the negative effect a comprehensive nuclear test ban could have on weapon reliability.
According to Deputy Assistant Energy Secretary Donald Kerr, test of nuclear devices in the past were necessary to show design defects and the unpredictability of weapons deployed to U.S. military forces.
Kerr said problems had occurred on more than a dozen occasions in the past 30 years "that have required tests for resolution . . . So one could argue that a problem on the average is going to occur once very three to five years."
In one instance a "serious problem" was found that would have affected more than half the weapons then in the U.S. stockpile. Kerr said yesterday that security classification prohibited him from describing exactly what was involved.
It was learned from sources, however, that a type of explosive system used to cause the nuclear reaction in several weapons systems had failed to operate as expected several years after the weapons had been put in the stockpile.
If the failure had not been discovered, sources said, the yields of the weapons would have been about one-tenth of what was expected.
In another instance, Kerr cited a weapon that was modified during the 1958-1961 moratorium on testing. Sources said the high explosive used to detonate the nuclear material was so sensitive it blew up while it was being machined for production, killing several workers.
A less sensitive explosive was substituted and the bomb production completed. The weapon - reportedly a 200-kiloton bomb - went into the stockpile without a test.
After the Soviets broke the moratorium on 1966 the modified bomb was tested and, according to Kerr, its yield was reduced "almost a hundredfold." Other sources said it exploded at about three kilotons.
Kerr gave his examples in a Sept. 29 letter to the House Armed Services subcommittee on intelligence and military application of nuclear energy.
The subcommittee has been sharply critical of Carter administration efforts to conclude a test ban treaty with the Soviet Union.
During its Aug. 14 hearings. Kerr had stated that testing was needed to make certain that weapons in the stockpile remained reliable.
That remark drew a letter of criticism from three respected scientists, two of whom formerly held key positions with Los Alamos Scientific Laboratory, the oldest of the nation's two nuclear weapons laboratories.
Former Los Alamos director Norris Bradbury and Dr. J. Carson Mark, the lab's former theoretical division leader, said there was no testing now of stockpiled weapons for reliability and none was needed in the future.
Kerr's letter was in response to that assertion.
Prior to the instances mentioned by Kerr in his letter, problems in the U.S. nuclear stockpile have been kept secrets.
Yesterday, however, officials close to the nuclear weapons program cited two other instances where tests had raised questions about the ability of stockpiled weapons to operate as intended.
In the 1960s, one source said, a particular safety device was put into nuclear warheads to prevent them from exploding accidentally. Several years after it went into deployed weapons, the device was tested. It was found to have deteriorated and was preventing the warheads from going off when they were supposed to.
In another case during the late 1960s test associated with an antibollistic missile system found that a deployed U.S. missile warhead would not work after being exposed to certain radiation. Changes in the nose cone of the missile had to be made to remedy the situation.
In a report released yesterday, the House Armed Services Committee recommended "no test ban be agreed to by the United States until a strategic arms limitation agreement has been ratified with the advice and comment of the Senate."
It also argued that any test ban agreement should "provide for weapons test of sufficients yield to assure to reliability of U.S. strategic and tactical nuclear weapons."
The Carter administration has been pushing a three-year test ban with only laboratory-size nuclear experiments to be allowed. The committee recommended tests of up to 10 kilotons (a yield equal to 10,000 tons of TNT) be permitted.
In a dissenting view, Rep. M. Robert Carr (D-Mich.) argued that "a mutual loss of reliability" in the nuclear stockpiles of the United States and Soviet Union might be a good thing in that it would make one nation worry about whether its weapons would work well enough to justify using them against another.