This is from the Minnesota Joint House-Senate Y2K Task Force Hearing on 7/14/1998, reported by Roleigh Martin, at:

http://ourworld.compuserve.com/homepages/roleigh_martin/news9807.htm#BM980719F >>>

Alliant [a power utility] reported about 170,000 embedded systems they are concerned with and investigating and of these about 3 percent are found to be noncompliant.

>>>

This is 170,000 embedded systems in just one utility and the indication is that there will be 5,400 non-compliant embedded systems, just for this one utility.

>>>

Bottom line: none of the utilities stated they had placed all of their orders yet for replacement equipment -- none of them are at that stage yet. Alliant mentioned they had approached some of their suppliers only to be told by some of them to "go away." They were quite concerned about the order problem.

>>>

Martin goes on to calculate the number of replacement systems that may be required, in terms of supplier capacity, and noted that the power companies themselves share that concern.

-- Rocky Knolls (rknolls@hotmail.com), July 22, 1998

Answers

This is very disheartening news. I spent the winter of '95 in Minneapolis, and temperatures dipped to -40F, without windchill factored in. It didn't reach 0 for weeks.

There's only one word for the Northern Plains without power: deadly.

Minnesota is, on the whole, a rather progressive state. That their utilities are facing these problems does not bode well for the nation.

-- Faith Weaver (faith-weaver@usa.net), July 22, 1998.

Hence my use of the term 'Digital Winter'! for Y2K problems.

-- Bob Barbour (r.barbour@waikato.ac.nz), July 23, 1998.

Alliant is a major utility in Iowa and Wisconsin as well. Unfortunately, it happens to be my utility and I was wondering how they were doing since they just completed a major merger of 3 utility companies. Good thing I planned for no electricity.

-- beckie (sunshine_horsees@yahoo.com), July 23, 1998.

Beckie, call Alliant's 800 number! They had a blip about Y2K in the last electric bill. You might be surprised at what they tell you. I bought a generator as a result!

-- Dave (dave22@concentric.net), July 23, 1998.

I don't want anyone to think I'm saying there's no problem, but you should bear in mind that "non-compliant" by no means automatically says that the noncompliant whatsit will stop working as of 1/1/2000 (unless they are using the term differently to most).

For example, consider a device that performs some vital process control and sends status and problem messages, time-stamped, to some central management system. These timestamps contain two-digit years.

Come 1/1/2000, this device will carry on working as if nothing has happened, except that the timestamp YY will be in an unanticipated state. It may be 00, or it may be :0 as the result of a common programming error. In either case, there's no immediate disaster provided the process control itself continues.

How much this matters depends on the behaviour of the upstream system. Furthermore, once you've established that the embedded whatsit is noncompliant and how it'll react to '00, you may not have to replace it. You could instead reprogram the management system appropriately so that the strange dates get correctly handled. This is a good thing to do if it's monitoring tens or hundreds of identically noncompliant whatsits; it saves both money and time.

Another example was encountered during a reported Y2K test of a power station (it had been shutdown for scheduled maintenance, and they set the clock forwards to late 1999 before restarting it). At just after the stroke of midnight 2000, an embedded system looking after the heatflow in a smokestack subtracted a time in 1999 from a time in 2000, got a period of nearly 100 years, thought that WWIII had just happened in its smokestack, and shut the plant down.

Again, that's noncompliant. The good news was that this problem was transient. A second attempted startup with the clocks already into 2000 was successful. So, here's a case where if the ten-second glitch was understood in advance, you could probably override the alarm for a few seconds and keep the plant going with nothing altered at all.

There ARE good reasons to worry about the power grid, but don't assume that "noncompliant" automatically means "will turn into a pumpkin at midnight".

The real problem will be the systems that are both noncompliant and not noticed until they fail. If it's true that some utilities don't have Y2K plans for their embedded systems at all, then there is definitely cause to worry!

-- Nigel Arnot (nra@maxwell.ph.kcl.ac.uk), July 23, 1998.

Nigel,

You quote of

<< There ARE good reasons to worry about the power grid, but don't assume that "noncompliant" automatically means "will turn into a pumpkin at midnight". >>

may very well be the best single statement I have seen made on this forum in months. Thank you for it.

Paul

-- Paul Neuhardt (neuhardt@ultranet.com), July 23, 1998.

Here is an EM from the y2k manager of a UK Power company, SWEB, he seems to think that embedded controls are not much of a problem: Subject: Re: Y2k & the Power Supply

Richard,

SWEB millennium information published on the internet (www.sweb.co.uk) explains the scope of our millennium project and identifies the types of systems which could be impacted. This information is in the process of being revised.

Briefly, any control system which contains embedded chips is potentially at risk from y2k impact. The actual impact depends upon the components used, the associated code, use of dates, etc. The consequences of impact depend upon the use of the control system and the nature of the failure or malfunction.

Embedded chips occur in many aspects of electricity generation, transmission and distribution. Although the concept of electricity generation and supply is the same, systems, manufacturers, technologies, etc, vary between sites, companies and countries. It is impossible therefore to give a general answer about impact.

As a critical resource it is vital that y2k impact on the provision of electricity is assessed and therefore concern is not misplaced. However, SWEB, like most other UK utilities, commenced y2k impact assessment in 1996. From our analysis the major area of impact in SWEB is in commercial IT systems. Although SWEB has a large operational asset base only a small subset of assets could be impacted. Of these a very small subset have been assessed to be at risk and appropriate actions can be taken prior to any impact. These would not have a direct impact on the provision of electricity.

Regards,

John Perriton

-- Richard Dale (rdale@figroup.co.uk), July 24, 1998.