[EPRI] INTERESTING! Rick Cowles' comment to Fred Swirbul on EPRI conference
Rick Cowles to Fred Swirbul: 'By your count, 75 *of the most
Y2k enlightened power companies* out of 9000 in the U.S. were at the
conference. I'm still preaching *awareness* to the other 8925.'
_________
'On Sat, 09 May 1998 04:43:49 GMT, Fred Swirbul <fswirbul@ix.netcom.com>
wrote:
>I just returned from a Y2K conference sponsored by the
>Electric Power Research Group (EPRI). Held in Dallas, Texas.
>Over 75 organizations were represented, mostly electric utilities.
>Other organizations are allowed to join EPRI's Y2K project
>(at $75K each). Oil and gas processors/refineries were the
>second largest group represented.
I would have liked to have been there, Fred - unfortunately, duty called at
other locations...
>Like anything else, there was both good news, and bad news.
>Good news/bad news. Most everyone had completed or was close to
>completing their initial assessments, and some preliminary testing
>has been accomplished by most organizations. Unfortunately, only
>about 10% of the organizations had completed significant portions
>of their testing.
Was PJM, NYPP, NERC, or any of the other regional power pool operators
present?
>Bad news/good news. Of those organizations that had completed
>significant testing, they were finding failure rates in the 10%
>range. The good news is that for embedded COMPONENTS, not a
>single "fatal" failure was found. Zip. Zero. Nada. I am calling
>anything with less computing power than a PLC (programable logic
>controller) an embedded component. Yes, the dates might be wrong,
>but the smart field transmitters still measured properly, the digital
>trend recorders still plotted trends, and the digital meters still
>displayed correct numbers (except for the date). Nothing at this
>level just froze up, so far.
This doesn't surprise me. As with Harlan, all along, it's been pretty
clear that problems at a 'chip level' are going to be in a teensy minority.
The problem with 'bad dates' at that level are probably not going to be
seen at that level, but as is typical with bad data, at a higher level in
the control system (say, level 2 or 2.5). I've never really expected to
see anything at ISA level 0 freezing anything up.
>It is starting to appear that it takes a fairly high level embedded
>SYSTEM to really screw up and lock up. A DCS (Digital Control
>System) or DAS (Data Aquisition System) can possibly fail in
>this manner. Even if 50% of all high level digital sytems have a
>Y2K problem (ie one of their many components is not Y2K compliant),
>it is starting to look like only one in ten will fail so bad as to
>trip a plant, whether it is a eletric plant or a refinery.
Taking the above at face value (for the sake of argument), one in 10
plants dropping out on a regional distribution system with, say, 200 plants
is 20 plants down simultaneously. That's way more than enough to
introduce enough voltage and frequency disturbances in the distribution
system to cause wild and crazy things to happen (ass/u/me that the regional
grid is totally Y2k ready with no problems to begin with).
>While the facts are just starting to come in, this seems to mean that
>most electric plants will only have a few systems that must be fixed
>before 12/31/99, so that they can still keep on producing power. IMHO,
>this does not appear to be insurmountable. If the Y2K problem is going
>to cause world wide hardships, it will happen with (most) of the lights on.
It's a matter of orders of magnitude, Fred. By your count, 75 *of the most
Y2k enlightened power companies* out of 9000 in the U.S. were at the
conference. I'm still preaching *awareness* to the other 8925. NERC is
just warming up to the issue. Regional system operators aren't in the game
yet. Please help me out - why should I feel any better today than I did
yesterday?
--
Rick Cowles (Public PGP key on request)
Now Shipping From AMAZON.COM: "Electric Utilities and Y2k" - The Book
euy2k.com
___
Subject:
Re: Embedded Systems Conference in Dallas
Date:
Sat, 09 May 1998 02:50:03 -0400
From:
rcowles@waterw.com (Rick Cowles)
Organization:
What's that?
Newsgroups:
comp.software.year-2000
References:
1 , 2 |
