I work for a rural county as an RPG programmer. We are preparing our plans for 12/31, and the question came up, "what are the Bonneville Power Administration (BPA) clocks set to?" We just contacted BPA, and the said "We operate on GPS time." They wouldn't clarify. My guess is that GPS time is Zulu (Greenwich Mean Time). Anybody know for sure? If the grid *is* on Zulu, then potential problems would hit at 4pm PST (where I am), or 7pm EST. This is crucial because we're starting our AS/400 backups at 1pm to be sure we're done by midnight. If the coach could turn to a pumpkin at 4pm we need to know soon!

Thanks for any help.

-- RPGman (tripix@olypen.com), December 02, 1999

Answers

OK. We just got confirmation. GPS time is 13 sec. ahead of GMT. As an aside, the father of an employee here works for the railroads, and tells us that railroads are also on GPS time. Boy we're gonna have to come in early on 12/31!

-- RPGman (tripix@olypen.com), December 02, 1999.

GPS = Global positioning satellite You can use the GPS system instead of good ol' WWV radio time to initiate a clock. So know you know WHERE they get their time, but not what they set it to, local or UTC.

-- Ken Seger (kenseger@earthlink.net), December 02, 1999.

BPA tells us the set their clocks to UTC.

-- RPGman (tripix@olypen.com), December 02, 1999.

The World Clock on the web translates between time zones. It is at: http://www.timeanddate.com/worldclock/ The USA Eastern Interconnection grid is tied into Canadian Hydro- Quebec which has a time zone in Halifax. So even if GMT does not determine a grid black-out of the North American continent, I believe the zone in Novia Scotia is the one to consider secondly as the "midnight" for the USA. However GMT is even earlier so I would say get your backups ahead of GMT midnight to be sure. l0ve, gl0ria

-- Gloria Watkins (watkins@dtc.net), December 02, 1999.

If most systems are set to Greewich Mean Time, TPTB who plan to watch for failures to happen at midnight 12/31/99 in New Zeland, Japan and other places that are hours ahead of GMT ARE GOING TO SEE NOTHING HAPPEN! They will gleefully proclaim their righteousness and the Pollys will rejoice..............until 7pm EST, (or a time zone near you) and then BANG!

-- Charli Claypool (claypool@belatlantic.net), December 02, 1999.

Here's what one article says:

http://www.capitolalert.com/news/old/capalert01_19990927.html

[snip]

The reaction could be influenced not only by Y2K problems developing in Asia, Europe and Africa, which will ring out 1999 ahead of the United States, but also in this country. A large chunk of the nation's electric grid, for example, operates on Greenwich Mean Time, meaning its computers will flip over to 01/01/00 at 7 p.m. Eastern time on New Year's Eve.

[snip]

-- Linkmeister (link@librarian.edu), December 02, 1999.

OK. I've wondered about this for a long time. Chips are made everywhere, right? If they have embedded date functions, don't they have to be universally set to something? Can't we really expect everything to go BLAMMO at the same time?

-- margie mason (mar3mike@aol.com), December 02, 1999.

Margie - that's the biggest problem with embeds - there isn't any "rule" about setting them to anything - particularly for those not cotrolled directly and immediately by outside "clocks" and computers - the damned things could be anything.....and that affects the quality of "testing" too.

Many, (we think, we hope, we expect) of these orphan controllers get "reset" to to a default date each time they start - so they "endlessly" keep beginning again at Jan 01, 1980 each time they get started each day like the old boot-from-a-floppy PC's - or some other equally wierd date.

----

It's almost impossible to say whether any given "set ahead" systems test -what few have been done, this is - has actually reset all embedded processors, most of them, none of thjem, or the critical ones, or only the ones that people knew about.....leaving a few "hidden" surprises even after testing....and of course those who have tested nothing or little will be expected to get even more surprises.

Like that steel plant in Australia that tried to test their system after remediation - only to see it gradually and completely go dead as various controllers and processors simply stopped.

---

Also, you've got to understand that even if the "grid" or any given processor is at GMT/GPS/or local - it will be reacting differently to different influences at different times....and will have "local" effects that are local time influenced - like client computers in a network.

-- Robert A. Cook, PE (Marietta, GA) (cook.r@csaatl.com), December 02, 1999.

Thanks for all the input folks. We *will* be getting up awful early 12/31 to start those AS/400 backups. "Get Real" asks a question in a thread above about Australia being our trip-wire. Anyone have any info on how the Aussie grid time is set? I was hoping to get a feel for how it's falling out by checking the tube early 12/31 to see if Sidney is dark.

-- RPGman (tripix@olypen.com), December 02, 1999.

I got the following response from a purported BPA employee. I have no reason to doubt what this man is saying. But now I'm more confused than ever: --- Jxxxxx, Hxxxxxx - TNC - AMPN-2 wrote: > > RP, > > I don't know who you have been talking to at BPA but I think there is a > misunderstanding. > > We do run off of GPS. However that is then SET to local Pacific Standard > Time (Pacific Daylight in the summer). I also don't understand the 13 second > time difference. I think that is another misunderstanding. > > Hxxxxxx Jxxxxxx > BPA ---

My reply... Mr. Jacobi:

Thank you for the information. I forwarded your email to the man at the courthouse who contacted BPA. BPA did tell us that their systems were set to "GPS time". Perhaps they misundertood our questionk, or we misunderstood their answer. We realy want to get this cleared up. We'll follow up on it tomorrow.

Thanks again, and God bless.

-- RPGman (tripix@olypen.com), December 02, 1999.

Basically, the "grid" doesn't run on any time, but what you have are EMS and SCADA systems used by utility operators to make transfers, etc. These systems are typically on local time from what I am told, with GPS time being used for accuracy, converted to local time. I could be wrong here, so if anyone in T&D has better info, please post.

Now, this means little in regards to power utilities and Y2K, IMHO, since all of the power plants I am aware of set the various Computer systems, control systems (DCS, etc) and lower level device date/times to local time. I have been in many plants, and never found any set to UTC, etc, it's always been local time.

Regards,

-- FactFinder (FactFinder@bzn.com), December 02, 1999.

So, theoreticly, the a computer that takes GMT & converts it to local would first 'see' a time as 11:59:59 12/03/99 and convert it to 06:59:59 12/03/99, right? Subtracting 5 hours for EST.

So, what will happen when the computer 'sees' 00:00:01 01/01/00 and tries to subtract 5 hours from it?

-- Arewyn (isitth@latealready.com), December 03, 1999.

Australia is upside down, so their clocks turn counterclockwise (as seen from where we're standing...) and their digital clocks run on the wrong side of the ac frequency input....

This means that they have actually been counting down to 01/01/2000 from above, and have already experienced "rollover" - they just couldn't tell because the tribes in 0000 didn't have electricity in the grid back in that future.....8<)

-- Robert A. Cook, PE (Marietta, GA) (cook.r@csaatl.com), December 03, 1999.