Uh, I thought it was 2, now all of a sudden it's 9.

Educate yourself:

Y2K shows N-plants didn't know equipment

Yomiuri Shimbun

Several Y2K-bug disruptions at nuclear plants across the nation have demonstrated that the electric power companies operating the facilities were not fully aware of the workings of their high-tech equipment and that their efforts to squash the millennium computer bug were unable to prevent human error.

Admittedly, none of the malfunctions adversely affected the safe operation of the nation's nuclear reactors. Still, the incidents highlighted the inadequacy of measures taken by the electric power firms to deal with the Y2K glitch.

Since Jan. 1, 10 minor malfunctions have been reported at nuclear power plants, nine of which were determined to be related to the Y2K bug.

"Fukushima No.1 and No. 2 nuclear power plants (in Fukushima Prefecture) and Kashiwazaki-Kariwa nuclear power plant (in Niigata Prefecture) have 3,707 built-in clocks. We confirmed that none of them has a problem," Teruaki Masumoto, managing director of Tokyo Electric Power Co. (TEPCO), said in September.

He made his remarks to members of the Atomic Energy Commission and the press after TEPCO conducted a Y2K countermeasure test at the Fukushima No. 1 nuclear power plant.

TEPCO officials were relieved to find that the end-of-the-year simulation they set up to test their equipment's Y2K compliance revealed no abnormalities.

However, at around 9 a.m. on Jan. 1, in the No. 1 reactor of Fukushima No. 2 nuclear power plant, an instrument for monitoring the location of 185 rods that control nuclear fission reactions indicated a malfunction had occurred.

The reactor was kept in operation for about three more hours before workers could confirm whether the control rods were using another computer.

The company discovered that some of the equipment's built-in clocks were on GMT, so the malfunction occurred when the program read the year 2000 as 1900 when the year ended at midnight GMT, or 9 a.m. Japan Standard Time.

The malfunction occurred in a piece of equipment installed by TEPCO in January last year and not in older machines, which are not equipped with built-in clocks.

Because workers did not know their equipment was set to GMT, they conducted Y2K tests using only Japan Standard Time.

An official in charge of the company's nuclear power plant technology said, "Though things have become more convenient, the machines now are like black boxes."

At Hokurioku Electric Power Co.'s Shiga nuclear power plant, a device for sending data about the status of nuclear reactors to the International Trade and Industry Ministry in the event of an emergency malfunctioned, and the transmission of data was temporarily hampered. The machine had passed inspection for Y2K glitches.

Workers who tested the machine were unaware that a program to increase the speed of data processing would misread the date.

There was also a series of incidents involving simple human error.

A TEPCO data storage program failed because workers in charge of repairs forgot to modify the program to make it Y2K-compliant.

A similar failure occurred at Tohoku Electric Power Co.'s Onagawa nuclear power plant in Miyagi Prefecture.

Though the International Trade and Industry Ministry's Natural Resources and Energy Agency had announced that all important computer systems, including those at nuclear power plants, were ready for Y2K, an official of the agency admitted these problems were significant.

Yoshinori Moriyama, chief of the agency's office of nuclear power plant operations, said, "We must regard it as a lesson for the future that failures to find problems occurred at more than one company."

Jun Sakurai, a technology critic, said, "If another problem had occurred while the location of the control rods was unknown, it might have taken too long to discover it, which could have had serious consequences."

"As the criticality accident and other difficulties have shown, new technology has become harder for the engineers to understand. An increasing number of engineers only follow the manual and do not know the insides of the machines," he said.

Copyright 1999 The Yomiuri Shimbun

-- Puddintame (achillesg@hotmail.com), January 05, 2000

Answers

Thanks for posting this Puddintame. I am not surprised by the number. On Dec. 31 (in the USA) we heard about one, and then an hour and a half later we heard of two....I had a feeling any 'news' would trickle out slowly.

Jun Sakurai, a technology critic, said, "If another problem had occurred while the location of the control rods was unknown, it might have taken too long to discover it, which could have had serious consequences."

We humans have been so incredibly blessed, there have been so many 'minor' glitches that could easily have come at the wrong time, or been made worse by human error.

-- Deborah (infowars@yahoo.com), January 05, 2000.

It'a incredulous to me that a reactor which is in a state where the exact position of control rod withdrawal is unknown was not manually scrammed. As a youngster in the nuclear Navy I was trained on and became a certified reactor operator. Of the many "commandments" I learned this type of event should be immediately be rectified by taking the appropriate manual measures to shut down the plant and not sit around guessing where the rods were - this is Three Mile Island mentality. Heads would fly if this were to happen at an NRC licensed site. If any of you current operators out there disagree please feel free to chastise me! But this one even makes common sense doesn't it?

-- Witch Doctor (bobmarley@hotmail.com), January 05, 2000.

Good thing they were all minor, readily detected, and easily corrected. And even better they do not appear to be Y2K related. All the facilities were checked and certified compliant in advance.

>"<

-- SH (squirrel@huntr.com), January 06, 2000.

An EXCELLENT article, forthright, and detailed, I doubt to see many like it.

An additional comment: the control rod software probably doesnt actually manipulate the control rods. What it does is MODEL the behaviour of the control rods/reactor and predict for the engineers how the reactor OUGHT to be behaving.

It is a critical piece of safety software, however as long as you know it's confused/misbehaving you can ignore its "answers"... Ah, there's th

-- Papa SMurf (thesmurf@ix.netcom.com), January 06, 2000.

One aspect of these Y2K failures is that when they failed, they became readily apparent. Alarms went off or displays otherwise gave indication of the error. These systems failed to a condition which drew attention to themselves.

What concerns me is the possibility that some systems out there were designed in a way that such failures would be masked. Forum comments about false alarms on oil pipelines just before rollover contrasted to last year's pipeline leak and explosion in Washington with a failed alarm are what I mean.

Right now, what is the potential that there have been Y2K related failures of embedded devices used in monitoring and control functions that are effectively "frozen" in their rollover state and will be mute or inactive when there is a change in operating conditions?

How much effort is going to be required to ferret such things out?

WW

-- Wildweasel (vtmldm@epix.net), January 06, 2000.