To: John Mansfield who wrote (427 ) 5/26/1998 3:54:00 PM From: John Mansfield Read Replies (2) | Respond to of 618
[HARLAN] 'Here's my current view of electric utilities.' 'In order to construct any kind of reliability model for the electric power infrastructure of USA/Canada, that could be queried to produce information regarding the probability of "service interruptions" to any group of customers, one would have to collect a large amount of information. Most of the critical information needed to drive this model, such it could provide information of useful accuracy is not yet available anywhere? Some is collected in electric utility databases -- one Y2K remediator told me he had a Lotus Notes database in which he was following the remediation/ repair status of 2,000 individual items. Major elements of the overall USA/Canada system are the nuclear, fossil fuel and hydroelectric generating plants, the transmission grids, and the distribution systems. There are about 108 nuclear reactors, hard to count number of fossil fuel plants about 7800 utility companies (large and small) of different kinds, 122,000 substations and switchyards etc. One can't really come to useful conclusions about how reliable the electric power supply will be to an consumer group without knowing a great deal about most of these system elements. These 7800 utility companies are all monitored and driven by a number of regulatory authorities of incompetent structure. I say "incompetent structure" because it is completely uncoordinated and has many huge gaps and overlaps. The Nuclear Regulatory Commission (NRC) for instance has the responsibility for making sure that nuclear reactors are safe, before permitting them to operate. There is _no regulatory authority charged with making sure that they will be safe and hence capable of operating. In essence they will perform their required function if they assume the attitude "I don't care whether you are safe or not, but if you can't jump through hoops to convince me you're safe, you won't run". A recent NRC letter, published here, indicates that they all must prove Y2K remediation by June 1999 or be subject to being shut down. Those having remaining loose ends at that time cannot continue to operate unless they present detailed information to the NRC indicating that they will very shortly tie up their loose ends. There are probably 25 - 40 billion "embedded systems" of all kinds in our global infrastructure and a few million in our electric utility infrastructure. A fraction of these are "date sensitive". Present testing and analysis seems to indicate that the simple microprocessor-implemented date-sensitive components will not significantly impact reliability of nuclear, hydroelectric or fossil fuel plants. The Y2K problem is then "thought" to reside in a few thousand "embedded systems" and computers at the complexity level of PLC (Programmable Logic Controller) and above. I understand that there were approximately 75 attendees at a May EPRI (Electric Power Research Institute) conference in Dallas. While that is a small number, information presented on news:comp.software.year-2000 indicates that the attendees represented about 70% of nuclear generation capacity and 45% of fossil fuel generating capacity. It is now said that nuclear generating stations supply about 18% of USA power, although for Chicago it is 70% and up to 50% for areas on the Eastern Seaboard. Utilities are doing a great deal of component testing, analysis, study and planning before they expend generating station outage time on remediation and testing. Most of the work is planned for normal "maintenance outages" rather than "forced outages" driven by Y2K remediation needs. May '98 EPRI conference attendees indicated that about 10% of their generating station systems had been scrubbed through and about 10% of "embedded systems" were discovered to have Y2K problems that require remediation. So, we are beginning to get a "feel" for the nature and scope of the problem. Even though we have thousands of problems, rather than millions, that doesn't mean that there isn't a huge problem. Remediation of any single "embedded system" may be quite complex. I have written a small article that discusses some of the difficulties with remediating "embedded systems". Embedded System Remediation: cpsr.org To start your reading on the subject of Y2K and electric utilities I recommend Rick Cowles and Roleigh Martin's web sites and _most _specifically the articles they have both written for the Westergaard Site. Rick Cowles euy2k.com Roleigh Martin ourworld.compuserve.com Westergaard y2ktimebomb.com (search on electric utilities) The Electric Utility WWW Resource List dei.gr --- I also refer you to my May 10th article "Synergistic Mitigation and Contingency Preparation" See: scotsystems.com and: ourworld.compuserve.com ___ Subject: Re: Michael Hyatt's _The Millennium Bug_ now shipping Date: 26 May 1998 13:02:14 EDT From: "Harlan Smith" <hwsmith.nowhere@cris.com> Organization: Paperless Newsgroups: comp.software.year-2000 References: 1