The scariest prognosis for the grid I have seen

greenspun.com : LUSENET : Electric Utilities and Y2K : One Thread

Rick and everybody,

I read an analysis today written by a man claiming to be a 30 year veteran COBOL/machine language systems programmer for IBM mainframes. His name is John Koehler and the piece can be found at http://www.gate.net/~koehler/y2k001.html#x002. He starts the essay with a disclaimer that it is only his personal opinion and that "it may turn out to be pure fiction".

There are quite a few doomsday prophets on the web, and it seems that some of them actually WANT the world to end as we know it. But this essay is particularly interesting because of the writer's credentials, if genuine, and his apparent depth of understanding of the problem and especially its effect on the Power Grid.

I urge you to read the piece and comment on the possibility of his scenario unfolding next year. It seems wildly radical to speculate that the grid could be down for 8 to 10 years. His reasoning is that the components of the grid are run by mainframe computers, and that those computers are broken, cannot ever be fixed, and will "no longer be of value" after January. He also states a pretty emphatic case for the embedded chip problem. He says "Embedded chips are everywhere, embedded in concrete, whereabouts largely unknown", and that if just 2% fail, along with all the mainframes, well, syanora civilization.

Another part of his rant involves nuclear plants. He thinks even if President Clinton decides by July 1st to shut down our 103 plants, there are plenty of nukes overseas that won't be, and they will inevitably melt down.

Mr Koehler's view is absolutely the worst worst-case scenario I have seen. Do you think we could experience anything like half of what he predicts?

-- Anonymous, May 24, 1999

Answers

Wayne, there is no specific data given to support any of the utility predictions or any of the other calamities mentioned. I do not agree that the article indicates a depth of understanding of the Y2K problem. I am not very tolerant of anyone who claims 100% failure of anything - or 100% non-failure, for that matter. For heaven's sake, there are lots of businesses nowadays who don't even use mainframes anymore. And NO ONE is going to decide on July 1st whether or not to close any nuclear generating plants. That's just the date the final plant surveys are due. Then those will be evaluated, more info obtained if needed, and any decisions will not be made until after August at the earliest.

Also, mainframe applications in utilities are found on the IT Business side of things, if they are there at all; work was begun on those first and is farther along than other areas. Electric generation does not stop if there are billing problems. (If it did, we'd have several outages right now! *grin*)

I pick the writer's own option of "pure fiction" on this one. Besides, Infomagic has this beat when it comes to scary predictions, and with more thoughtfulness put into the prognostications. *wink*

-- Anonymous, May 24, 1999


Yep. Don't know about Infomagic and don't care too, but Bonnie's right about the rest.

-- Anonymous, May 25, 1999

Based upon my technical expertise, I draw the following conclusion:

Bonnie, you go girl!

-- Anonymous, May 25, 1999


Wayne and Bonnie, I have not read the article yet, Bonnie, and I hate to burst your bubble, however, as a purchasing agent and manager for construction of large plants, I know we could be in for a very very serious problem, such as that discussed by you, Wayne.

This has been one of my main nightmarish fears. Even when we built these things in the US, when the economy and factories were good, and most things we needed were manufactured in the US, it sometimes took years of lead time to get many of the items built which were necessary to build these plants. This includes computer control systems.

Now many of these items are no longer manufactured in the US and we must import them. We have become an "information" society. Just as you cannot build a plant with a paid insurance claim, you can't build a plant with a valid purchase order either.

If we have problems which takes down and damages major components or even whole plants, I have no problem with an estimate of 8-10 years to rebuild/replace them, or even longer, if ever.

I remember once, a few years ago, on a side trip after the oil crash, while working for NASA, I had to order some component parts for the NASA shuttle. The original manufacturer had gone out of business. We had to find a supplier to gear up a factory, purchase tools and equipment then, re-design and fabricate these components from scratch. These are something similar to an automobile gas cap in purpose, size and appearance, called a flight cap, used to cover hypergolic fuel filler connections. We needed about 75 of them, of various configurations. They had origninally cost about $300 each. It took 4 months of work to just find a company who would even do it, then 18-24 months for them to complete the job and they cost generally from $25,000 to $46,000 each.

Now, I admit that this is an extreme example, however, it has happened to me on several occasions for other items. General procurement time for major components in large plants currently is 1 to 3 years. That is without any high demand from others competing for similar items.

Look at the FAA, working for years on a new computer system, which still doesn't work. Look at the IRS, $4 billion spent before they gave up on their new system.

So, without even reading the article, unfortunately, I can support part of the prediction, the time line, whether the other aspects are true or not.

-- Anonymous, May 25, 1999


Bonnie,

I believe that mainframes may pose the biggest risk for the power industry because I think that they may be holding the largest amount hidden mission-critical complexity.

You said: "Electric generation does not stop if there are billing problems." I believe that we may be puting this statement to the test soon.

-- Anonymous, May 25, 1999



xBob, no need to apologize, I like all the different viewpoints I can get! If the article in mention had put forth specific reasons behind the assertion that we would be without electric generation for months, besides loss of mainframe capability, then there might have been more to debate.

The basis of your concerns (valid ones, I think, in ways I'll get to further along here) depends upon that "If we have problems which takes down and damages major components or even whole plants.." From a U.S. centric view only, I think the evidence indicates an extremely low likelihood of any major plant damage if there are some Y2K failures in utilities.

That said, there is another aspect of the 2000 rollover that puts your concerns center stage, and it's another breed of problem altogether. I think as far as any government *can* make anything clear (smile), ours has already been pretty straightforward about the possibility of terrorist acts superimposed upon the change of century. Whether those acts might take the form of computer viruses or bombs planted at utilities or elsewhere, I don't think it's naive to factor in the possibility of such things occurring. In the case of terrorism, the repair of any physical damage to plants or substations could well be hampered by the procurement problems you cite.

Ironically, there is an upside to the Year 2000 computer problem, as far as the possibility of terrorist actions goes. Instead of everybody having one big unconcerned global party, almost all key infrastructure areas will now have a lot of people at work and intensely focused on looking for and dealing with any potential trouble. It hasn't been mentioned too much, but it has been put forth by government agencies that heightened attention to security matters and procedures would not be amiss in the coming months. (Or at any time for that matter, but we Americans are still pretty casual in maintaining an "it can't happen to us" mindset.)

As for the IRS, well, that's off topic here, but I personally wouldn't bet a plug nickel on their viability in 2000. (Or is that wishful thinking? *grin*) The death-by-a-thousand-cuts scenario is one that transcends the normal focus of this forum on how ready the utilities are for the rollover. Actually, I think that's a good part of the friction which appears here sometimes. Utility workers are looking at their individual generating capability and others are factoring in the many adjunct potential problems for which preparation remains wise (in my view) regardless of power on or power off, or for which longer-term effects may be out of a utility's control.

A person out of work because of Y2K failures in their business (even a temporary lay off which I've seen happen several times because of computer problems stopping factory production) can still be sitting in the dark, not because of no generation, but because they can't afford to pay their electric bill. There is a difference between possibilities and probabilities, any insurance actuary could tell us all about that. I consider the probability of many potential scenarios to be very low, but that doesn't mean I don't consider the possibilities aren't valid ones. IF any of the most worst-case problems do come up in the Y2K lottery of life, then it's my position that all bets are off, preparation or no preparation. It's the larger middle ground where prudent risk management will be of benefit.

Reporter, you bring up another view of the broad picture. There is the possibility that any prolonged billing/metering or other IT management process problems could seriously erode a utility's viability over time. However, I see that as having the potential to cause a take over/buy out by a more stable utility than to cause generation per se to stop altogether because the utility business end folds up. On the probability side, it's think there's a good chance that electricity prices will rise in the future, deregulation or no. The costs of Y2K repairs and the costs of any failures will eventually be factored in and passed on to consumers. They always are, whatever the business.

Whatever an individual's personal assessment of possibilities versus probabilities is, it's my own belief that we are entering into a period of substantially increased risks to infrastructure and global stability. I take those risks seriously, although I do not and cannot know the precise impact of them beforehand. We'll be putting a lot of things to the test soon. I hope we get at least a grade of C-plus! Getting either an A or an F are other low probabilities in my personal book of assessments.

-- Anonymous, May 25, 1999


The business of the grid is mostly run on mainframe computers. The grid itsself is not. COBOL knowledge on mainframes just as Foxpro knowledge on PCs does not imply ANY expertise on embedded systems. Hard, time dependant realtime systems programming is a different world entirely. I wouldn't touch anyone's accounting software and I wouldn't want any business systems programmer anywhere near my safety critical realtime systems.

Jim

-- Anonymous, May 27, 1999


It's been a while since I checked in with this forum, but I'm glad I spotted this thread.

Question: Lots of companies are working very hard on fuel cells to power homes and businesses. The technology seems to be leapfrogging (either that or pronouncements of same are designed to raise those stock values).

In a worst case scenario: fuel cells to the rescue?

(Also off topic here, but I just spent a grueling two days transcribing the complete first hour of oral testimony--including the question and answer session--at Monday's Senate hearing on Y2k preparedness. I'm offering it to the public domain. I plan to do the same with the second hour. So, if you'd like to read the first hour, or have interest in the second, visit the following thread:

http://www.greenspun.com/bboard/q-and-a-fetch-msg.tcl?msg_id=000sS3

Yours truly,

(Hot-link impaired)

:)

-- Anonymous, May 27, 1999


Along the same lines as above:

Electricity in a Box

BPA to Install High-Tech Fuel Cell Generators in Northwest Homes to Produce

Clean, Efficient Electricity

PORTLAND, Ore., May 27 /PRNewswire/ -- The Bonneville Power Administration today announced it has signed a contract to purchase 110 fuel cells from Northwest Power Systems (NPS) of Bend. If testing of the new technology is successful, the region will lead a high-tech revolution in the electric power industry.

"Our goal is to adapt clean, efficient fuel cell generators to small-scale consumers and commercial applications," said Jack Robertson, deputy administrator of BPA. "In just a few years, we could see these energy boxes distributed as widely as the home computer.

"BPA will take delivery of the first proton exchange membrane (PEM) fuel cell systems this fall. Northwest utilities will test the 3-kilowatt units for use in homes. After the first ten 'alpha' units are installed and operated, NPS will make any necessary adjustments and build 100 more 'beta' test units," said NPS President Alan Guggenheim.

BPA will work with local utilities to place the beta units in the homes of interested customers. Trials of prototypes by BPA over the last two years were successful and the agency decided to move ahead with a full-scale test, Robertson said.

"Although fuel cells will not replace large central generating plants, they can help meet load growth and provide clean, efficient electricity in homes," Robertson said. "Over time, the efficiency of these and other types of distributed generation will make them the choice of consumers."

The power units are 85-percent efficient when waste heat is recovered for space and water heating. Conventional generators are less than 35 percent efficient in the use of fuel. The cost of producing the beta units is about $30,000 each, but Guggenheim expects that the price will drop to under $10,000 per unit when they become commercially available in the year 2002.

Fuel cells produce electricity from natural gas and other fuels such as methanol, ethanol or propane. NPS has designed a processor that chemically removes hydrogen from the fuel. The proton exchange membrane then strips electrons from the hydrogen, thereby generating electricity. The process virtually eliminates the carbon monoxide, carbon dioxide and other harmful gases emitted by combustion engines.

Initially the fuel cells are expected to be particularly useful in remote locations as backup generators, and in applications where reliable power is particularly important. BPA supports the development of clean, efficient energy technologies as part of its conservation and renewable energy program.

Northwest Power Systems is a subsidiary of IDACORP Technologies, Inc., in Boise, Idaho. The company aims to develop automated electric power generating systems ranging from 1 kilowatt to 10 kilowatts in power output by next year.

-- Anonymous, May 27, 1999


Moderation questions? read the FAQ