GAO Report on Nuclear Power Plant Security
The GAO reports a mixed bag. Security improvements are progressing on paper, we're just not sure if that reflects reality. The Nuclear Regulatory Commission is "not visiting the plants to obtain details about the plans and view how the plans interface with the plants’ physical layout," nor is it "requesting, and the facilities are generally not submitting for review, the documents and studies supporting the draft security plans." GAO-04-1064T, 9/14/04 [3-4]. This repeats an administration pattern: leaving our security in the hands of private corporations; even heavily regulated industries are being given a free hand on security measures. Isn't this an area where oversight is not just appropriate, but necessary?
The most egregious example may be this:
The agency is planning to require the use of an adversary force trained in terrorist tactics, as we recommended in our September 2003 report. However, NRC is considering the use of a force provided by a company that the nuclear power industry selected; this company provides guards for about half the facilities to be tested. This relationship with the industry raises questions about the force’s independence. GAO-04-1064T, 9/14/04 Seriously, why bother? If the same company that is providing security for the plants is going to orchestrate the practice security ops, will we be able to get reliable or useful performance information? Given the track record of the company, the answer is a resounding "no":
[T]his company was recently involved in a controversy over similar tests. During a June 2003 DOE force-on-force exercise at a nuclear site in Oak Ridge, Tennessee, security guards working for this company received uncharacteristically high scores. A subsequent investigation by DOE’s Office of the Inspector General indicated that the guards might have cheated on the test and perhaps on many other tests at Oak Ridge, dating back to the mid-1980s. It was alleged that the guards had studied plans for the simulated attacks before they were carried out, had disabled the laser sensors they wore during tests to determine when they were “shot” by mock enemies, arranged trucks and other obstacles to help foil simulated attacks, created special, nonstandard plans to help them perform better on tests, and put more guards on duty at the time of the tests than would normally have been present. GAO-04-1064T, 9/14/04 [12-13]The report highlights are here [PDF].