A 55-gallon drum of nuclear waste, buried in a salt shaft 2,150 feet under the New Mexico desert, violently erupted late on Feb. 14 and spewed mounds of radioactive white foam.

The flowing mass, looking like whipped cream but laced with plutonium, went airborne, traveled up a ventilation duct to the surface and delivered low-level radiation doses to 21 workers.

The accident contaminated the nation's only dump for nuclear weapons waste — previously a focus of pride for the Energy Department — and gave the nation's elite ranks of nuclear chemists a mystery they still cannot unravel.

Six months after the accident, the exact chemical reaction that caused the drum to burst is still not understood. Indeed, the Energy Department has been unable to precisely identify the chemical composition of the waste in the drum, a serious error in a handling process that requires careful documentation and approval of every substance packaged for a nuclear dump.

The job of identifying the waste that is treated and prepared for burial will grow even more difficult in the years ahead when the Energy Department hopes to treat even more highly radioactive wastes now stored at nuclear processing sites across the country and transform them into glass that will be buried at future high-level dumps.

The accident at the facility near Carlsbad, N.M., known as the Waste Isolation Pilot Plant, or WIPP, is likely to cause at least an 18-month shutdown and possibly a closure that could last several years. Waste shipments have already backed up at nuclear cleanup projects across the country, which even before the accident were years behind schedule.

A preliminary Energy Department investigation found more than 30 safety lapses at the plant, including technical shortcomings and failures in the overall approach to safety. Only nine days before the radiation release, a giant salt-hauling truck caught fire underground and burned for hours before anybody discovered it.

The report found that "degradation of key safety management programs and safety culture resulted in the release of radioactive material from the underground to the environment."

The 15-year-old plant, operated by a partnership led by San Francisco-based URS Corp., "does not have an effective nuclear safety program," the investigation found.

The accident raises tough questions about the Energy Department's ability to safely manage the nation's stockpiles of deadly nuclear waste, a job that is already decades behind schedule and facing serious technical challenges.

"The accident was a horrific comedy of errors," said James Conca, a scientific advisor and expert on the WIPP. "This was the flagship of the Energy Department, the most successful program it had. The ramifications of this are going to be huge. Heads will roll."

There is no official estimate of the cost of the accident, but outside experts and a Times analysis indicate it could approach $1 billion, based on the WIPP's annual budget; the need to decontaminate the facility; upgrades to safety that officials already have identified; and delays over the next decade in the nuclear weapons cleanup program.

The WIPP was designed to place waste from nuclear weapons production into ancient salt deposits, which would eventually collapse and embed the radioactivity for at least 10,000 years. The dump was dug much like a conventional salt mine, but with a maze of rooms for the waste. It handles low- and medium-level radioactive materials known as transuranic waste, the artificial elements — mainly plutonium — created in the production of nuclear weapons. Until the Valentine's Day disaster, it had been operating without significant problems for 15 years.

The plant's ventilation and filtration system was supposed to have prevented any of the radioactive material from reaching the environment. But investigators discovered that the Energy Department never required the ventilation system to meet nuclear safety standards. When monitors detected radiation, dampers were supposed to route the ventilation air into filters to prevent any radioactivity from reaching the surface, but the dampers leaked and thousands of cubic feet of air bypassed filters.

Luckily, the accident occurred when nobody was working in the mine itself. But the emergency response moved in slow motion.

The first high-radiation alarm sounded at 11:14 p.m. When control room managers tried to find the responsible on-call radiation control expert, they couldn't find the person, according to the investigation report. By morning, workers were attempting to change filters. Not until 9:34 a.m. did managers order 150 or so workers on the surface of the site to move to a safe location, about 10 hours after the first alarm sounded. It took 13 hours for managers to staff an emergency operation center.

The radiation doses the workers received during the hours after the accident were a small fraction of the allowable occupational limits and the workers should have no health impacts, Energy Department officials said.

Although WIPP operating procedures were faulty, the dump itself did not cause the accident. The steel drum was packaged at the Los Alamos National Laboratory in New Mexico. The drum principally contained nitrate salts, a byproduct of the chemical process that extracts plutonium, used in the triggers of hydrogen bombs.

Investigators believe that some chemical or packaging change was made at Los Alamos, and they are looking at whether that change was ever approved by senior laboratory chemists. A team of experts from WIPP may also have missed the change.

The investigators are looking at a variety of materials that may have been added to the drum, including lead, tungsten, acid and even kitty litter as possible factors in the explosion.