<I tend to put more credence in an anonymous SI poster cutting and pasting a random e-mail transmission from someone who heard it from his neighbor, who had talked to someone in the bar who once worked at a nuke plant, etc.>
Subscribe to usenet group: comp.software.year-2000 - Then decide who is more credible. The guys in the front office, or the guys actually doing the work.
Check out House Testimony on Embedded Systems. They use real examples here. This part deals with NRC:
RADIOACTIVE MATERIALS WASTE TREND ANALYSIS
Radioactive material waste tanks are monitored and some are controlled by automated sensors and other devices. They all work on date sensitive trend analysis. What will happen to trend analysis when there is perceived to be a 99-year span between two measurements? Who is responsible?
ATOMIC/NUCLEAR SITES(example from the U.K.)
Software on nuclear sites is subject to stringent quality controls. However, experienced software industry professionals already grappling with the Year 2000 have expressed doubts about how reliable these design-based reassurances are. Hard and fast test data to back up these assurances has not been provided.
The first area of concern is the radiation exposure system. The program for the control of radiation exposure is called ALARA (As Low As Reasonably Achievable). Nuclear facility personnel wear dosimetry devices that measure the amount of whole body exposure that the employee receives while in the plant. These dosimetry devices are analyzed on a regular basis and the data (exposure amounts) are maintained on a computer system that control personnel access. To meet Nuclear Regulatory Commission (NRC) and The Institute of Nuclear Power Operations (INPO) regulations the exposure amounts are monitored on a daily, weekly, monthly, quarterly, and yearly basis.
Second, a "Training Records Tracking System" computer controls access and actual work assignments to ensure that the Reactor Operators, Second Assistant/Auxiliary Operators, Maintenance Technicians, Radiation Protection Personnel, and Plant Management employees have completed the required initial and requalification training for their work assignments.
The third area of concern related to the Year 2000 is the computer system that tracks various plant commitments for hardware and operation procedure improvements.
When considering the impact of the Year 2000, the following questions arise.
1. Will all plant personnel risk exceeding radiation exposure limits because the ALARA computer system is inoperative?
1. Will unqualified employees be allowed access to the plant and work assignments because the Training Tracking computer system is inoperable?
1. Will plant personnel be at risk because of expired respiratory protection qualifications?
1. How will the Department of Energy (DOE) control, track and inventory uranium 235/238, plutonium, tritium, or americium with Year 2000 problems?
1. Will plant commitments be delayed or not completed on time because the commitment tracking computer system is inoperative?
1. Will unqualified operations personnel be operating the reactor in the control room without the required initial classroom training, on-the-job training, qualification card sign off or requalification training?
1. Will personnel be wearing respirators with expired qualifications (e.g. annual physical examination, medical screening, annual radiation protection requalification training, mask fit process)?
1. Will there be any clearance requirements for the computer professionals to correct the Year 2000 problems at DOE facilities? Is there enough time for the computer professionals to obtain the DOE (Top Secret "Q") security clearance and still have the time to fix the computer systems prior to January 1, 2000?
1. Will the maintenance schedules on plant hardware be carried out properly if computer based records fail?
house.gov
__________________________________________________________________
JIM: I sincerely hope NRC is in a better condition than DOD.
Defense Department's Global Command and Control System flunked a Year 2000 rollover test during an Interoperability Exercise in August 1997 ..... In 10 of the 28 demos, either the software expired or the machines froze ..... One system crashed when it tried to cross over to 2000 and when it was rebooted, all user accounts were gone.
cv.nrao.edu Sept 22 '97
Control systems in power plants, T&D, and offices all depend on "embedded software" that might fail at the turn of the century. This "Year 2000 Problem" is quite different from the well-publicized Y2K problem in utility business software applications running on mainframe computers. The impact of embedded systems failure, while not known, could potentially be severe and widespread.
epri.com
Electric Power Research Institute |
