By Robert O'Harrow Jr.
Washington Post Staff Writer
Tuesday, June 23, 2009
The Department of Homeland Security spent three years pushing for a costly nuclear detection system that does not work as billed, while neglecting to upgrade existing equipment that could have helped improve security, according to a new U.S. Government Accountability Office review and an interview with one of the authors.
The report is the sharpest critique to date of one of the Bush administration's marquee national security programs, and it marks the culmination of several years of struggle by the GAO to determine whether the detectors worked as well as government officials claimed when first promoting the $1.2 billion project in 2006.
New technical tests last year found that the machines, known as Advanced Spectroscopic Portal monitors, detected potential nuclear threats slightly better under limited circumstances than existing gear at border crossings. But the new monitors cost $822,000 each, more than twice as much as the equipment now in use, the GAO said.
"The report shows the pitfalls of pushing to deploy technology before it is ready," said Gene Aloise, director of natural resources and environment for the GAO. "At least three years have been lost."
The report follows a sometimes rancorous debate over the merits of the advanced spectroscopic program, which was embraced by the Department of Homeland Security's Domestic Nuclear Detection Office (DNDO) as a way to sharply reduce false alarms triggered in existing equipment by benign forms of radiation in cat litter, ceramics and other non-threatening materials. Homeland Security officials have said those false alarms create delays at borders.
In recent testimony to the Senate Homeland Security and Governmental Affairs Committee -- which has pressed questions about the machines for two years -- Homeland Security Secretary Janet Napolitano made clear that she would not request funding this year for the machines. That's in part because the DNDO has tens of millions of dollars left over from the previous two years' budgets.
"And the reason is because we were not persuaded that . . . the capacity of the technology that we needed was actually there," she said.
Contracts to spend up to $1.2 billion on 1,400 devices were first announced in July 2006. Then-Secretary Michael Chertoff said that they were a "vital priority" of the administration and that they were ready to be deployed.
The next month, GAO auditors alleged that the machines could not perform as well as the department claimed and questioned the department's cost estimates of $377,000 per machine.
In response, Congress mandated that the secretary personally certify that the detectors would bring significant improvement to existing machines before moving forward with a full-scale deployment.
The validity of subsequent testing for that certification was then questioned by GAO auditors, who said that DNDO "used biased test methods that enhanced the apparent performance of the ASPs." That bias included allowing contractors to adjust their machines after preliminary runs, enabling them to appear to perform better.
The new GAO report focuses on another round of testing showing that the new devices perform better than existing equipment when the radioactive material is lightly shielded in something such as a cargo container. That advantage decreased when more substantial shielding such as lead was used, as likely would be the case if a terrorist were trying to import the material, the GAO said.
The GAO report also found difficulties in integrating the equipment at border crossings. In one instance, officials at Customs and Border Protection "suspended field validation of the system after 2 weeks because of serious performance problems that may require software revisions," the report said. The ASPs were sending more false alarms than the older equipment.
And DNDO has not moved forward on potential techniques to improve the sensitivity of existing radiation detection equipment. If those machines performed better, Congress might be less likely to approve the costlier ASPs, the report said.
"It is disappointing to learn that the next-generation ASP system provides only a slight improvement in detection performance," said Sen. Joseph Lieberman (I-Conn.), chairman of the homeland security committee. "That's why Secretary Napolitano is right to hold back on ASP certification and to seriously consider a new approach."