From http://www.tva.gov/news/today/da111998.htm#2 comes this good news.

I can't believe there wasn't coverage in any of the regional news media about this.

************************************

It's Y2K at Gallatin & Chickamauga - and the Lights Are Still On

TVA's Gallatin Fossil Plant and Chickamauga Dam rang in New Year 2000 several months early to get an idea how things will go when the new millennium rolls around for real.

Both sites completed Year 2000 unit-startup tests last month, and both continue to run using 21st-century dates. The goal of the tests was to ensure that all equipment with date-related functions will operate properly when the date 01/01/00 arrives.

Startup conditions were chosen as the best test of plant equipment. The reason is that changes occur in all plant systems at startup, offering a good look at how equipment will operate during those changes. At Gallatin, the coal-handling system also was included in the test.

Before startup, all clocks at each site were set to let them naturally progress to Jan. 1, 2000. The leap-year date of Feb. 29, 2000, also was tested. The consensus of many in the utility industry, including the Electric Power Research Institute, is that testing these two dates yields an accurate indication of equipment's Y2K compliance.

All four power-generating units at Chickamauga and units 1-3 at Gallatin were successfully tested. (Due to a planned outage, Gallatin's unit 4 will be tested at a later date.) The clocks at both sites will be left to run in the year 2000 until early January 1999.

The Y2K team that conducted the test included representatives from Fossil & Hydro Engineering, F&H Operations, Information Services, Telecommunications and Transmission/Power Supply.

-- Anonymous, November 23, 1998

Answers

David, this is truly good news. However, I noticed just below the article you posted here was another one about major fixes scheduled for Thanksgiving weekend. Perhaps they are waiting to see how this goes before releasing anything to the press? I believe some of these systems would be of critical importance to continued operation. At least they are implementing them early enough to get any problems straightened out, but that's one heck of an upgrade load to schedule all at once. I bet the IT people will be turning the air blue that weekend. *smile* I wish them all good luck.

"Major Systems To Be Upgraded for Y2K Over Thanksgiving Weekend

Procurement announces that the systems listed below will be unavailable Nov. 26-29, Thanksgiving weekend. These systems are being upgraded by Information Services to make them Year 2000 compliant.

Acceptable Suppliers List (ASL) Advance Shipping Notice/Barcode System Automated Release processing (ARL) Automated Sourcing Project (ASP) Business Partner Access (BPA) Contract Information System (CIS) E-mail Notification System Federal Procurement Data System (FPDS) Freight Transportation Management System (FTMS) Integrated TVA Materials Systems (ITMS) Classification ITMS Accounts Payable/Purchase Order (AP/PO) ITMS Fossil Fuels (FF) MAMS Bar Coding MAMS EMPAC Interface Materials Acquisition Tracking System (MATS) Materials Management Systems (MAMS) Inventory (including Ad Hoc Reporting) Office Supply Ordering System (OSOS) Plant Maintenance Procurement Data Warehouse (PDW) Supplier Performance Rating System (SPRS)

Also, MPAC and EMPAC which are work-management systems used primarily by Fossil & Hydro and TVA Nuclear will not be receiving any information from or sending any information to any of the above systems during this period, says Procurement.

Upgrades to these systems must be made before Jan. 1, 1999, in accordance with TVA's Year 2000 Project. The Thanksgiving holiday weekend was selected for these upgrades to reduce the impact on day-to-day TVA business and to allow time to assess the results of the upgrades before the end of the calendar year. "

-- Anonymous, November 24, 1998

Here is a BIG question that I have. It is my understanding that a large part of our worries regarding the electric utilities is EMBEDDED CHIPS, which I understand might fail or cause undesirable things to happen at Year 2000 BUT cannot have their dates set forward. (Specifically, it is my understanding that the only recourse with embedded chips is to 1) locate via inventory; 2) determine whether it has a Y2K problem, usually by contacting the manufacturer; 3) REPLACE if it has a Y2K problem, because once again its code cannot be modified.) When a utility says it has set the clocks forward to Year 2000 HOW CAN IT BE ACCOUNTING FOR THE EMBEDDED CHIPS??

-- Anonymous, November 24, 1998

Jack, If the control system is running off of one clock and the chips have their own, I would think that the whole thing would have gone down already due to the discrepency between the components that think it's still 1999 and the ones that think it's 2000. Somewhere, sometime, a date stamped chip will tell the system (or a technician) that the high pressure injector pump, or the exciter generator, or some such thing was a year past it's recommended service time for preventive maintenance which can cause a trip-out. Could also be that the remediation code has an "include" routine that performs an interpretive function for two digit dates until the date chips are made compliant. In any case, they've reached the full testing phase a year before the changeover which is enough time to work out the bugs. Wish more utilities were that far along! All success Keith

-- Anonymous, November 24, 1998

Jack:

May I add this to Keith's remarks.

In the November 23 post by PC Week: "An Embedded Dilemma: Embedded System woes are overhyped but too serious to ignore," the story challenges the number of embedded chips presently in use, and subdivides them into two groups:

"Dump Systems" which are usually not vulnerable to Y2K, and the less numerous, more complex "True Embedded Systems" that contain some sort of software code. In theory, the article states, these systems can be upgraded if necessary (instead of being replaced). The story acknowledges the challenges of such an endeavor, citing that source codes have to be located, and that those codes could be written in languages no longer used.

How this might relate to the type of embedded systems utilized by the electrical industry, I cannot answer. But evidently, not all non-compliant systems have to be replaced. Some can be upgraded.

As a sidebar, this report dovetails with Harlan Smith's recent postulating that embedded systems are the real issue, not embedded chips. Smith claims too much time is spent looking for the 1 in a 100 chips that might fail and that we should be examining the larger and more complex embedded systems.

Microcontrollers that are used for a single basic function are not likely to have the capacity to shut down a system, come 1/1/2000, Smith argues.

This stance may in fact affirm the Gartner Group's much maligned claim that embedded systems pose less of a threat than previously thought. In reading their position on the embedded technologies issue, the "1 in 100,000 failure rate" figure refers to microcontrollers and not embedded systems. As mentioned before in other posting to this site, Gartner actually pegged embedded system failure rates at "up to 35%" for some industries.

Rick may be able to confirm or refute this figure in terms of its application to the electrical industry, but it looks like there may be fewer systems to consider, however a higher rate of Y2K-induced failures within that number. Hopefully additional testing by the industry will clarify this emerging position on embedded technologies.

-- Anonymous, November 24, 1998

Sorry, folks. That was supposed to be "Dumb systems," not "Dump systems." What do you expect from a dyslexic who can't type well either.

-- Anonymous, November 24, 1998

Somewhere lately I read that some systems tested as described (roll-over naturally into 01/01/2000 with no problem) have experienced problems in the first reboot subsequent to rollover. I suppose TVA will find out.

-- Anonymous, November 27, 1998