WASHINGTON - There is a budget crisis, but the truth is we're still planning to spend tens of billions of dollars to eliminate plutonium from thousands of dismantled, surplus nuclear weapons built during the Cold War.

Not to worry. We also are spending hundreds of billions on building newer nuclear warheads and bombs, and 21st-century submarines, bombers and intercontinental ballistic missiles to keep more than 1,000 nuclear weapons at the ready.

One thing that building and destroying the weapons have in common: Their cost overruns are way beyond original estimates.

One such program - to dispose of 34 metric tons of weapons-grade plutonium - got full exposure last Wednesday at a Senate Appropriations subcommittee hearing. First estimated by the Energy Department in 2002 to cost $3.8 billion over 20 years, the project is now projected to cost more than $31 billion.

Radioactive materials such as plutonium decay very slowly and continue to emit potentially harmful particles long after the warhead or bomb is dismantled.

In 2000, the United States and Russia signed an agreement in which each country would convert 34 metric tons of surplus weapons-grade plutonium into a form no longer useful for weaponry. They also would have to approve the other side's method, and the International Atomic Energy Agency would serve as monitor.

For both countries, the idea was to mix the plutonium with uranium to make mixed-oxide (MOX) fuel for use in reactors that produce electrical power.

While Russia chose to use its MOX in what are known as fast reactors, the United States initially chose using MOX in light-water reactors. The overall U.S. approach required constructing a new MOX facility, another to disassemble the nuclear weapon triggers (or "pits") and a facility to handle radioactive waste.

It also required finding a buyer for the reactor fuel produced.

In 2002, the Energy Department estimated the total cost of the MOX project over 20 years would be $3.8 billion. The department's facility at Savannah River Site in South Carolina was chosen as the site for the plutonium conversion.

As Sen. Lindsay O. Graham, R-S.C., noted at the hearing: "South Carolina said we will take this highly toxic 34 metric tons of weapons-grade plutonium . . . into our state with the condition that it will go out of our state and the federal government would honor the commitment."

Concerned because the South Carolina facility was already stuck with waste from its tritium-production plant, Graham recalled that he put a deadline into law that "requires 1 ton of plutonium to be processed through MOX or shipped out of the state of South Carolina by 2016 or pay my state $100 million a year for five years."

The most recent estimate is that the MOX project, between 40 and 60 percent completed, would not go into operation before 2019.

But rocketing costs last year got the program slowed down while a task force was set up to assess MOX alternatives.

The original $3.8 billion had grown to $4.8 billion by 2008 and to $7.7 billion by 2012. Also, annual operating costs of more than $500 million were said to be underestimated, driving total cost to more than $31 billion.

At last Wednesday's hearing, retired Air Force Lt. Gen. Frank Klotz - just 10 days on the job as the Energy Department's new undersecretary for national security and head of the National Nuclear Security Administration (NNSA), told the senators, "The current MOX approach must be critically examined alongside costs of other potential options to complete the plutonium disposition mission."

The report of the study begun last year, he said, laid out four MOX alternatives. They would be reviewed, he said, and a way forward would be decided in 12 to 15 months.

Meanwhile, Klotz said, work on the MOX project, on which some $5 billion has been spent, will continue through Sept. 30.

Sen. Lamar Alexander, R-Tenn., noted how a similar situation existed with another NNSA program - the construction of a uranium facility at Oak Ridge, Tenn. to replace one built in the 1940s for the first atomic bomb.

That Oak Ridge project's cost also ballooned, Alexander said. It started "at $100 million, and next thing we know it's $2 billion. ... Then it's $5 billion, and then it's $6 billion, then it might be $10 billion."

He said a "red team" appointed from several laboratories studied the project and in 90 days came back with "what looks like a perfectly obvious central solution."

The growing costs related to building one huge, high-security building would be far higher than building two facilities - one with lower security requirements - that would cost less than $6.5 billion, the report said.

Wednesday's hearing ended with Klotz agreeing to brief Alexander and Sen. Dianne Feinstein, D-Calif., on a reassessment of the costs of the current MOX program so it could continue.

Meanwhile, across the Capitol, a House Armed Services subcommittee tentatively approved a provision that would require the administration to sharply increase production of new plutonium pits to 50 as a war reserve beginning in 2026. The administration has already put $2.5 billion into the plutonium pit facility at Los Alamos National Laboratory.

It turns out 20 pits annually and is expected to reach 30 annually by 2026. The House panel's language would require greater pit production than what Los Alamos could provide and four years ago was projected to cost $4.5 billion.

All this begs the question of why so many new plutonium pits are needed since current ones will last more than 60 years, and it's expected that warheads will be reduced in the coming decades.

With all these complaints about wasted spending in domestic programs that help the less fortunate, why aren't the excessive costs of nuclear weapons activities being debated in Congress?

Walter Pincus is a columnist for The Washington Post.