FAIRBANKS -- The 26 rockets in underground silos, built at Fort Greely to defend against a missile attack on the United States, are each topped with a device designed to stop an incoming ballistic missile with blunt force.
There are other complex elements in the missile defense system, but the business end of the enterprise is the "Exoatmospheric Kill Vehicle," or EKV, which is intended to collide head-on with an incoming missile outside the atmosphere.
The "kill vehicle" itself is 2 feet wide, less than 5 feet high and weighs about 140 pounds. Onboard sensors and guidance controls allow course adjustments at speeds of thousands of miles per hour. It has to be aimed perfectly.
But a new report this week from the Pentagon office charged with testing and evaluating weapons systems questions the reliability of the EKV.
Test missiles have never been launched from Fort Greely, about 100 miles southeast of Fairbanks in Alaska's Interior, because of the threat that falling rocket stages might strike a populated area, but there have been several tests of the equipment from other sites on the Pacific Rim.
The EKV has not performed as well as it should, according to Michael Gilmore, director of Operational Tests and Evaluation for the defense agency.
"The flight test failures that have occurred during the past three years raise questions regarding the robustness of the EKV's design," his report said.
It said additional testing is required, but that the Missile Defense Agency should "consider whether to redesign the EKV using a rigorous systems engineering process to assure its design is robust against failure."
The Ground-based Midcourse Defense element of the nation's missile defense system, with major rocket facilities at Fort Greely, has "demonstrated a partial capability to defend the U.S. homeland from small numbers of simple intermediate or intercontinental missiles," the review said.
There are 26 missiles at Fort Greely, and the Pentagon plans to add 14 by 2017.
There are two versions of the EKV. A newer version called the CE-II had a successful "non-intercept flight test" last year, but the first version, the CE-I, had an unsuccessful flight test.
The agency is planning another test in March, using a CE-II that has been modified to correct a problem from an earlier flight.
Last July, there was an unsuccessful test in which a CE-I did not separate from the interceptor.
A 2012 Government Accountability Office report said 12 of the CE-II devices had been delivered, but two "attempts to verify its capability in flight have failed -- the first because of a quality issue and the second because of a failure in the guidance system."
"The guidance system failure has resulted in design changes and the risk of further changes on the program will remain until flight testing is complete," the GAO said.
The cost of the CE-II test failures could exceed $1.2 billion, the GAO said. Retrofitting each of those devices would cost about $18 million per interceptor, according to the agency.
At that point the failure of the flight tests meant the capability of the dozen CE-II EKVs that had been delivered had not been demonstrated, the GAO said.
In 2012, the director of the missile defense project said that after a 2010 test failure critical components of the CE-II EKV were redesigned and more stringent manufacturing and component tests standards were set. He said the implementation of those standards led to testing delays.
The Missile Defense Advocacy Alliance, a group that lobbies for missile defense projects, portrayed the comments by Gilmore as positive, saying "the cost of U.S. investment and the technology growth learned in fielding this first and second generation GBIs (ground based interceptors) is validated and continues to be of tremendous value and insurance to the American people."