Why would power failures be of short duration?

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread

The power industry has been working on remediating the plants, {nuke or fossil} for say a year or more, and still have 9 months to work on them. If they fail, why is it assurmed by so many that they will stay down only a short time? If they could not be fixed in 2 years or more, why do we expect that they will they be fixed in a few weeks after they fail? It would seem to me that if they go down they will stay down a long time. Are we depending on a fix on failure policy? Will they be more motivated if they are working in candle light? I am very worried that if we have black outs they may be a way of life for a long time. Please explain to me how I am wrong.

Bill in South Carolina

-- Bill Solorzano (notaclue@webtv.net), March 16, 1999


You are probably not very wrong. To me it seems that this "fix-on- failure" attitude is just another way of ignoring the issue.

Good luck to us all...

-- Marcus Sokolowski (Marcus.Sokolowski@imim.uu.se), March 16, 1999.

The reason that 'fixes' will be prompt is usually stated thusly: when the system goes down, we will be able for the first time to identify the points of failure, and go in with the necessary replacements. Although not a line problem, the mechanics of fixing it will be similar to a line problem. And it won't be in candlelight, because we have generators, and the ability to let juice flow to necessary sectors while the failed components are identified and replaced.

-- Spidey (in@jam.com), March 16, 1999.

You're not wrong. "Short Duration" is the sure-fire spin designed to keep people as sheeple. Eagles UP!

-- Buffalo Bob (buff@hal.com), March 16, 1999.

I don't have a full understanding of that, either, but here is one analyst's rationale for choosing 72 hours as his projection for the initial blackout: http://www.y2ktimebomb.com/PP/RC/dm9901.htm (sorry, can't seem to make it a hotlink) He was more concerned about power shortages later in the year, due to a long-term decline in capacity caused by Y2K.


-- Margaret (janssm@aol.com), March 16, 1999.

the thinking, flawed or not, is that this problem in data handling, data switching, and product control is equivalent to an ice or wind storm, where the physical infrastructure takes a hit and is fixed in a "short" time. What the analogy leaves oout is that, as has been said, "Fix on Failure" is NOT like cruising the streets looking for the limb or downed line. You are cruising lines of code, hoping you find the one or two that coughed up the dump, and hoping that you can either rewrite the routine, and have everything run until you hit another couple lines of code.

Also, you may be cruising the system schematics, until you find the bad controller, or chip, and hopefully you have a "compliant" analog to replace it with. If all you do is pull out an XTG-453-MK1.01 and replace it with an XTG-453-MK1.01 you have NOT solved the problem, soomething the "FOF" people haven't taken into consideration.

The "FOF" croud clearly expects that the individual enterprises have identified ALL of the controllers and embedded chips, and have ON HAND compliant versions as replacements for ALL of them. As I said above, if a subsystem is bad (ie, it dies while trying to run in 2000) an identical replacement will likely die also.


-- Chuck, a night driver (reinzoo@en.com), March 16, 1999.

Having followed debates on utilities closely here and on Cowles' forum, I suspect Chuck has his finger on the dike. If embedded systems prove "no problem", outages can be controlled and limited, even if nasty (anything > 24 hours is nasty from this dog's vantage point). If it turns out that too little real assess and tests were done (there are still some embedded systems experts who are scared to death), diagnosis on failure and unavailability of replacements could worsen situation dramatically (you also get chicken and egg with supplying systems from a supply chain that is "down", partly due to the same util outages).

As Chuck has been saying lately, our problem remains (and will remain) lack of good data from those doing the stuff to know whether we can relax our guard or not.

By and large, I accept Cowles' evaluation, for the unapologetic reason that I trust his overall integrity: we're probably not going to see grid meltdown, just lots and lots of "crap" happening of all kinds for a significant period of time (up to a year, with situation improving over time broadly). That's bad enough for me to plan for the worst, hope for the best. Too many "probablies" (and I'm talking Rick here, not the PR of the utilities, which I both understand and loathe: Y2K is not a normal situation and demands more than the usual CYA).

-- BigDog (BigDog@duffer.com), March 16, 1999.

Bill, This is a good post.

Electrical interuptions are not the best times to be fixing anything.

What do folks do during interuptions? Rush to work? Work with what?

What about other interuptions? Do they not create serious distractions that interfere with ones focus and abilities?

-- Watchful (seethesea@msn.com), March 16, 1999.

Spidey wrote: "The reason that 'fixes' will be prompt is usually stated thusly: when the system goes down, we will be able for the first time to identify the points of failure, and go in with the necessary replacements."

Just where are these "necessary replacements" coming from, if these parts have not previously identified, ordered in a Y2K-compliant configuration, and kept in inventory? The fix-on failure argument is predicated on the belief that the replacements for damaged parts are available and compliant. I doubt that this will be the case.

-- sparks (wireless@home.com), March 16, 1999.

There is another issue related to continued electricity that I don't understand. Even if the electric companies fix everything with the power lines, substations and all their computers, how can they continue to generate if the nuke plants and trains that carry the coal are not working.

I have read articles saying the railroads are in trouble. It's no seceret that nuclear power plants are not ready. Electricity is generated from nuclear power plants and coal fired plants, primarily.

Soooooo, how can the electric companies deliver power when there is nothing (or not much) to deliver? If there is a small amounts of power being generated by hydro and whatever other sources they have, it seems to me it will go to hospitals, police, water plants first. Households will be last on the list. And that's giving that the electric companies will be compliant.

I am preparing for no electricity.

If the embedded chip turns out to be as bad as some fear. The power companies will take years to fix everything. How are they going to get the new embedded chips to replace the old ones?

I'm preparing for 10 years with out power because I can't firgure out how they can fix it.

-- monique (me@home.here), March 16, 1999.

Ever wonder why we don't seem to have power industry experts (apart from our esteemed nuclear experts) posting here on the forum?

From my viewpoint, unless you have personal expertise in the power industry, the question boils down to whom do you believe?

Those who are saying expect months and/or years of no power and a return to 1900's living have absolutely no expertise in the power industry and generally make statements based on fear.

On the other hand, engineers who have genuine expertise in the power industry and who have no reason to lie (i.e. Rick Cowles and Dick Mills) are telling us that any outages will likely last for a few days with the potential for rationing or shortages later due to capacity problems at plants.

I may have missed it, but I haven't read a single rebuttal of Rick Cowle's interview with Drew Parkhill claiming that Cowle's is too optimistic. Until someone with a greater level of expertise comes along and says otherwise, I'm going with that.

I may be prepared to go for a few months without power just in case, but I still think the greatest impact Y2K will have is in the long-term economy.

-- David (David@BankPacman.com), March 16, 1999.

i tend to agree with u however with all the spins they are all making contingency plans and we are not supposed to......with the oil deliveries and rail problems and postal problems ect....they expect a three day fix.......honsety is the best policy however in the case of y2k its kinda a silly request ...or so it seems...

-- vicki (myfivekids@aol.com), March 16, 1999.

Vicki, 5 children? God bless you! As with you, my prep is much more extensive than would otherwise be because of my children (2 at this point). While I don't trust the government or utilities to speak the truth, Cowle's is not a spinmaster. Check out CBN's website for the link to his interview with Drew. It's pretty long, but well worth the read.

My understanding of the three days is not so much to fix anything. Rather, it would take that long to kick in the manual workarounds to keep power flowing while the fixes are being made. Granted that's an oversimplification of the issues involved. Since I don't know what I'm talking about, I'll defer to Mr. Cowle's for the proper explanation.

-- David (David@BankPacman.com), March 16, 1999.

Here's the link to the Cowle's interview in case you're interested:

http://www.cbn.org/y2k/cowles .asp

-- David (David@BankPacman.com), March 16, 1999.

It's only guesses now anyway, and will be until they try a real test of the actual distribution system and two or more power plants both on line at the same time with dates set ahead.

Most likely: a little up, then down a long time (3-6 days), shut down (or trip off line - try to find problem), try to start again - run a while, load changes -> trip off again, fix again (or reset manually), staty up a little longer, duty electrician gets tired from six days continuous work -> makes mistake - trip off again -> reset system -> stay up a little longer, trip off, try again, stay up a little longer than before, trip off, try again, stayu up a little longer.....run out of parts, stay off a while, ....

-- Robert A. Cook, P.E. (Kennesaw, GA) (cook.r@csaatl.com), March 16, 1999.

And related questions- how much of your collection of household electric/electronic appliances will survive a couple of months of low voltage/spikes/dirty power/surges that are likely to be the norm if our resident P.E. is right? And who will pay to replace them?

-- (li'ldog@ontheporch.com), March 16, 1999.

Going back to Chuck's comment, while the programmer is scrolling down through endless lines of code looking for THE one that is bad, he will probably be "playing" with it and "fixing" some of the good lines as he goes. I have just the tinest little bit of programming in my educational background (just enough to know how difficult it can be) and have learned that fixing code can be VERY iffy at best.

-- linda (smitmom@hotmail.com), March 16, 1999.

The biggest problem with "Fix on failure" is the dependency upon eletrical power. If the power companies are counting on fixing their problems in this way, ALL other companies that depend on electricity will not be able to do anything until the power companies complete their fixes. These companies will not be able to start ANY of their systems to even attempt to find out what went wrong, let alone fix it if there is no power.

So, how many months (or years) will it be before the electricity flows stable enough for most business to even BEGIN their "fix on failure" repairs?

-- Russell (Oh Boy@y2k.com), March 16, 1999.

I don't know if theres any weight to this. When I was talking to someone in emergency government in my area,I asked why they say 72 hours possible no power. He stated that is how long it would take the power company to island itself.

Maybe someone here knows if thats just hot air or a possibility?

-- maji (majiWI@yahoo.com), March 16, 1999.

Nobody brings this up, but I have another question. In reading early writings on the power stuff (Cowles and others), I noticed it mentioned that one of the things that can actually cause a localized failure is when there is more demand than there is power, something about it taking out transformers... you see I don't know what I'm talking about but I do remember reading this. If trains and coal, nuke plants etc. are all at issue, meaning we might have LESS power, and if the power keeps going on and off, isn't it possible that everytime it went on there would be the chance that lots of local transformers would blow due to the demand so overriding the availability? I'm not worried about it really, I'm just curious as I never did hear a follow up to this tidbit. I'm preparing for no power, because if I'm wrong, I'll just have extra candles and extra propane, and if I'm right, I'll be thanking God for my foresight.


-- PJ Gaenir (fire@firedocs.com), March 16, 1999.

PJ - yes, the transients and power durges that are a likely by-product of manual or temporary controls could blow things - what would happen under that circumstance would be very local - as far as damage goes, only the one transformer would be shorted out or cause a fire. However, the feedback and secondary effects - like the single transformer that blew in San Francisco can trip other units both upstream and downstream form the blown unit. Worse, if fire happens, that section must beimmediately deenergized to avoid subsequeent damage. Immediate isolation could be difficult if control remotely is haphazard.

Remember, in that case, the failure was at only one point, but ending up killing power to the whole penisula. That's why I believe that the surges and spikes inherent in damage recovery and trips will also do damage toelctronics - and greatly increase the problems troubleshooting to "fix on failure".

There is a second point about isolating (islanding) units from the grid that few address - if a unit is on the grid, the vast load already out there means that any transient is stabilized - even a whole building tripping off (or starting up) is a fraction of a small percent compared to the grid. But if a power plant is isolated, the effect of the same change in demand is magnified - so loads and feedback is made more difficult for the turbine controller to manage.

-- Robert A. Cook, P.E. (Kennesaw, GA) (cook.r@csaatl.com), March 16, 1999.

Moderation questions? read the FAQ