I've seen a couple references now to a major test being conducted on 4/9 of the "grid".

However, I have no information about the scope of the event, other that that it is supposed to be "grid-wide".

Anyone have further information/insight about the a) scope of the tests being conducted (IT/Non-IT systems, which ones)? b) who are the players involved? c) area of the tests (regional/grid/interconnect-grids)? d) independent confirmation of the scheduled date?

Thanks, Bob.

-- Anonymous, January 18, 1999

Answers

Bob, there is some info about this on page 45 of the recent NERC report to the DOE at: http://www.nerc.com/y2k

(You need an Adobe Acrobat reader to access this, however.)

Here's the first paragraph: "Initial planning of the April 9, 1999 industry drill has begun. A document describing the scope, objectives, and reporting requirements for this drill will be available by the end of January. This drill will focus on operating the system with limited voice and data communications. It requires placing personnel at key operating facilities and communicating by backup systems."

-- Anonymous, January 18, 1999

Bob, Another old programers "shortcut" besides a 2 digit year was to use the code 99/99/99 in the date field to indicate another operation should take place (IE. End of processing, delete record, obsolete record etc.) The reason April 9th is significant is that many applications also use julian dates for processing (xx day of the xx year.) Well April 9th is the 99th day of the 99th year or 9999 in computer language. So computers that have julian dates in addition to "9" codes, could be affected. Also remember computers do not use the "/" in their computing they are just for our ease of reading the info. So many think that the 9999 julian date could trigger some of the 99/99/99 responses and disrupt valid data by treating it as invalid or whatever the "9" code indicated.

-- Anonymous, January 19, 1999

Right.

Thanks, folks.

Bob.

-- Anonymous, January 19, 1999

In my opinion NERC has chosen 4/99 and 9/99 as so called "test" dates so that if there is a system failure they can suggest to the public that is was due to the tests being conducted. In other words the systems are "under control" because they failed during a "test" which failure was "anticipated" or due to "human error" during testing. Otherwise why pick these key programming dates?

-- Anonymous, January 28, 1999

What is the purpose of a drill? To test your response to either a "worst-case scenario" or a "most-likely" scenario. To stretch your resources to the boundry of their designed operating parameter and beyond. What is the expected benefit? To 1. Verify that you can respond to your intended limits. 2. To see if you can respond beyond expectations. 3. To identify strengths and weaknesses, share ideas/strengths with others so they benefit, to take action to cure weaknesses. Did you ever consider that maybe, just maybe the reason for the drill dates isn't sinister, but logical. Why mobilize your resourses and "create" an emergency for a fabricated drill? Why not have all your resources mobilized when there is an alleged potential for an actual incident? Would you rather the utilities have a full mobilization and drill on St. Patricks day and have everyone at home in bed on 4/9 & 9/9?

-- Anonymous, January 28, 1999

Ok you raise a good point Chicken but you are presupposing that there will be some big mobilization effort as part of this "test" or that they are going to actually do something to probe the systems when in fact they are just having time pass on these critical dates, seeing if there is any glitch and calling it a "test" regardless of the outcome. They could do that without characterizing it as a test. The problem is that if they don't call it a test it is an out of control system if it fails. It gives them plausible deniability to avoid panic which is probably a good thing in the end. I just don't like being fooled. If they are going to test why don't they just push the clocks up to 1/1/2000 (if that's even possible on embedded systems). That would be a test.

-- Anonymous, January 29, 1999