Northern State Power reported in its 10Q filed August 13, 1999 that it has finished 99 percent of its Year 2000 project with respect to both its mission critical and non mission critical systems.

It also reported that would take 20 percent of its Year 2000 budget to finish. (It has spent $19.1 million to accomplish the 99% but will require an additional $5.3 million to finish the remaining 1%.)

What does it mean when it says it is 99% done? How much of this 20% of its budget will it spend on the 1% of its mission critical systems?

-------------------------- From NSP's 10Q filed 8/13/99, as reported on www.yardeni.com: "This information is designated as a "Year 2000 Readiness Disclosure." NSP is incurring significant costs to modify or replace existing technology, including computer software, for uninterrupted operation in the year 2000 and beyond as discussed in NSP's 1998 Form 10-K.

NSP is on schedule for completion of its Y2K project.

o On June 30, 1999, 99 percent of both NSP's mission-critical and non-critical systems and processes were Y2K ready.

o On June 30, 1999, NSP filed its contingency plans as required by the North American Electric Reliability Council. NSP's contingency plans are comprehensive and include the following actions: the establishment of back-up or alternative data and voice communications, increasing generation reserves, coordination with government agencies, increased staffing levels during Y2K critical time periods and conducting readiness drills.

o By Dec. 31, 1999, NSP expects to have completed implementation and testing of all applications.

Since the start of the Y2K project in 1996 through June 30, 1999, NSP has spent approximately $19.1 million for Y2K efforts, which (except for a portion deferred for approved rate recovery) has been expensed as incurred. The additional development and remediation costs necessary for NSP to prepare for Y2K is estimated to be approximately $5.3 million"

-- Anonymous, October 01, 1999

Answers

Almost all large projects approach their completion asymptotically. It is the nature of things.

dave

-- Anonymous, October 01, 1999

These numbers just point out the problem we have had all along in trying to monitor the progress of all large business systems. NSP has given us the budget figures. They say they will spend another 5.3 million which would be a little over 21% of an original $24.4 million budget. But *was* there an original 24.4 budget? Or has this budget just kept rising from some earlier, lower, number? Because in that case, $5.3 million more than the current $19.1 million is an *increase* of 28%, not 21%. So the numbers look like they are being massaged, or played with, to come out looking a bit better. Does anyone really think they originally set a budget of $24.4 million and are still working from that number?

Then too, there is 99% done with only 1% to go. Something isn't right here, obviously. Is the 99% the amount of lines of original code that have been gone through? Is the 1% remaining the most difficult, most intransigent, lines of code to be corrected? Will it take many months to finish the final 1%, because that's what the budget suggests. Will it really matter if the last 1% is not finished? A lot of smoke and disconnect is going on here, but we have become used to that by now. In fact that seems to be the common mode in communicating progress from government or business.

If we focus only on the 1% "work" remaining, this looks great. If we focus on an "additional budget" of 28% more money to finish, it doesn't look great at all. I wonder who is putting the face on these numbers, and if that face is designed for public confidence rather than accurate accounting progress?

-- Anonymous, October 02, 1999

Snyder, several months ago I posted a portion of a letter I'd received from a friend who works for a major international consulting company. (You'd easily recognize the company name if I was free to give it.) He'd written me that in his opinion the wrong questions were being asked when it came to percentages. He went on to give a simplistic example which I'll change a little to fit your post. A company could have 100 systems, 99 with a few 100,000 lines of code each, and one with a few _million_ lines of code to be looked at. Thus the company could with all honesty state that they were 99% finished even if the biggest job was still ahead of them or only partially done.

This has always been one of the problems with percentage reporting, and may have something to do with your puzzlement. As Gordon has written, there are quite a few potential reasons for the NSP statement, or a combination of any of these things may account for the seeming inconsistency. I've also run across a utility which stated in an earlier 10Q that they would never declare themselves 100% complete in any Y2K area as they considered their Y2K project to be an ongoing effort to ensure that nothing had been missed. I personally interpreted that as a legal maneuver (and a pretty good one, at that) but it could as well have been an honest recognition that you just don't get 100% error free status on any computer system, Y2K or no.

There are so many inconsistencies in the Y2K status reports, I don't think we're ever going to know exactly what was meant. (If, in fact, anything exact actually _was_ meant in the first place.) It all boils down to waiting to see what happens in 2000. And I believe there's much less time left for any actual work to be accomplished between now and January 1 than most people realize. My husband and the other people I know who work in IT are all agreed that historically next to nothing of any import gets done between Thanksgiving and Christmas/New Years, other than an urgent rushing around to get the many and varied end-of-year reports finished.

Even last year NERC skipped the monthly survey report for December. As I see it, there's only about 30 some potentially productive work days left before Thanksgiving arrives. I think if a company isn't ready now, it's a pipe dream to think the next three months will make any real difference. As Tom Benjamin said not too long ago, it's time to sit back, put our feet up, and watch. Our government may still be going on about all that can be accomplished before year's end, but for me, it's Y2K Ready-Or-Not, right NOW.

-- Anonymous, October 02, 1999

Hi Bonnie and Gordon -

I agree completely with your comments. What I don't understand is that NSP has run full-page ads in the Fargo (N.D.) paper advising people there is no greater likelihood of a power outage on January 1st than any other day of the year. They started doing this months ago, along with other reassuring pronouncements. It indicated in June that it had not found any problems are all in its nuclear plant in Monticello Minnesota, even though NSP's Monticello plant was subsequently listed as a plant with a problem with its "core physics" on the NRC report.

I don't mind if they have the usual problems in fixing things. I do have a serious problem, in a place where it gets as cold as it does here, for them to give reassurances that discourage people from caring for themselves, when they still have 20 percent of their budget to spend.

Leon Kappleman made an excellent point in something I read recently. He said that some problems are certainly going to occur. If all the authorities have attempted to assure people that nothing will happen than their credibility will be severely diminished when some things do happen. Then, people won't know who believe since they may not feel they can believe government and business.

Leon made a point, that I agree with, that leadership would be far better off preparing people for some glitches than not. I certainly think that applies with regard to individual utilities. The Senate report is excellent at doing exactly that. Unfortunately, the one person in the national media who actually perhaps looked at it got it wrong, and all the other national media seems to be rehashing his/her story.

Sigh...SSDD - Same day, different S**** (..."stuff" - what did you think i was going to say?)

Snyder

-- Anonymous, October 02, 1999

oops....guess thats "Same stuff, different day"

-- Anonymous, October 02, 1999

Bonnie,

Y2K Ready-Or-Not, right NOW, you said, and I agree. Cory Hamasaki just recently stated that his best information from the IBM big iron, 390 types, was not very good. About 50,000 running the major government and business needs, worldwide. Cory thinks there may be, possibly, a few hundred that are fixed and ready. The rest, anywhere from close to ready, to hardly fixed at all. This comes from his personal knowledge and programmers who talk with him about this matter. This means problems, big problems, crash and burn type problems, and when they can be fixed at all, it will take a long, long time. Time enough to bankrupt a lot of those users. I really don't enjoy this sort of gloomy forecast at all, any more than the recent Floyd hurricane forecasts, but I have to live with the best information I can put together, and go from there. That was the way I ran my flights, and that is how I run most of my daily life.

As to statements from the utilities that things are really going to be no worse than any other January day, in any other year, I find that to be irresponsible. Even if they want to play a "we're OK" game, they could at least admit to their own supplier vulnerability and suggest that anyone with a fireplace stock up on more wood, or anyone with a kerosene heater store some extra fuel. Things like that. But their apparent unconcern for the welfare of their many individual customers is unconscionable, in my opinion. They are going to stick some people in a real bind. We have a couple posters right here from within the industry that show even less concern. They say they are not going to do anything different, no extra food even, because there is no probability of electric failure. This ignores the rest of the infrastructure entirely. They don't see, can't see, any connections beyond the generator lines from their plant to your house.

-- Anonymous, October 02, 1999

In the grasping at straws category, you could argue that 20% of the budget might be unspent after acheiving Y2K compliance. This money would be needed for retaining expert consultants, maintaing increased stocks of fuel and basic materials, money spent on implementing an IT freeze, money to be sepnt on extra personnel for contingency operations, etc. There could also be a lag between money spent and the accounts payable ledger. Not that I buy any of the aforementioned. Anyone see the latest Yardeni "Experts Poll" (not good news)

http://www.cio.com/info/releases/093099_y2kpoll.html

don't know if any of this is generalizable to electricity but its another one of those SSDD articles, slipping deadlines, lack of implemented contingency plans, unverefied vendor compliance statements ... and optimism abounding (at least from an IT management standpoint)

As regards stating full compliance, think about it how could anyone do this. They have just sent 30 million lines of code to the Fix-It Co. in New Dehli and gotten it back and running with an "all done" stamp and a poor translator trying to tell them how wonderful everything went. 30 million possible points of failure most of which are probably fixed but any one of which might under the right circumstances present a real problem. Now I'd be happy to hear about their efforts to fix this problem, but under duress of possible litigation do I expect them to say they're could never be a problem. Fact is no one knows if the whole program will throw a fit when it finds it didn't get a new calender as a present for the New Year or smile benignly and place a "00" in the right upper hand corner of the printout with a reminder to fix when convenient. Even if I were a cobol programmer, you can't review 30 million lines of code, and if you could you might miss that a particular subroutine causes a stack overflow or other bizaroma type error. Not to mention replacement of thousands of chips in all manner of configurations which in type testing worked great and which got along great with their old non- compliant chip buddies. Hopefully automated systems don't much care about the calendar, but aside from a company telling me their working on it I doubt I'd be much more impressed to hear about compliance percentages.

-- Anonymous, October 02, 1999

Percentages and Budgets: What Do They Really Tell Us?

-- Anonymous, October 02, 1999

I can see that it's time for a real academic (I.e. home-school mom) to set y'all straight on math. ;) Back to story problems... "Joey wants to drive from San Francisco to North Carolina to help with the flood victims. He has saved his money for the last three months because he knew he wanted to minister to others. He has $1500.00, and the trip is about 3000 miles. Joey has already brought his sleeping bag and a jar of peanut butter with him, so he won't need to stop for food. He expects to have $1000.00 left when he arrives in N.C. After three days, he has traveled 1200 miles. How much money does he have left?"

Follow the money!

-- Anonymous, October 03, 1999

He has $1500.00, and the trip is about 3000 miles. Joey has already brought his sleeping bag and a jar of peanut butter with him, so he won't need to stop for food. He expects to have $1000.00 left when he arrives in N.C. After three days, he has traveled 1200 miles. How much money does he have left?

This is tough. Math majors, you know, are notoriously bad at simple arithmetic. :-)

Since room and board are excluded as factors, fuel purchases would seem to be the expense under consideration. He expects to spend $500 during the trip of 3,000 miles ($1,500 total - $1,000 left). According to expectations, he would spend an average of about 17 cents (16.66667) per mile ($500 / 3,000 miles). So, according to expectations, after travelling 1,200 miles, he would have spent roughly $200 (16.66667 cents * 1,200 miles); so, he would have $1,300 left ($1,500 - $200), $300 of which (16.66667 cents * 1,800 miles) would be spent on the rest of the travelling, according to expectations.

This kind of analogy, unfortunately, has flaws so serious with respect to Y2K that it is more misleading than it is anything else. The chief flaw is that expenditure of budget is assumed to be done at a constant rate: this assumption is simply false. Testing, for instance, can cost much more than any other aspect of remediation -- can cost more, not necessarily does cost more -- especially if a "time-machine" is leased. Another flaw is the "expectation" of how much money would be spent: such expectations often fall short of reality, though they might also be shown to have been overblown, too.

Actually, the trip analogy, if done realistically without assuming a constant rate of expenditure, can do quite nicely. The price of gasoline, for example, can vary considerably from one locale to the next. And fuel efficiency (miles/kilometers per gallon/liter) can vary considerably depending on many factors: terrain, traffic, etc. So, the assumption of a constant rate of expenditure is false, as in Y2K remediation. There might be other factors, too. Maybe Joey blew a tire and discovered that the spare was flat, so he had to spend $60 on a new tire in New Mexico. Or maybe he picked up a grateful hitch-hiker in northern Texas who gave him $10 for gasoline.

So, the percentage-of-budget-spent figures are pretty much useless for indicating, or determining, how far along a Y2K project is. The general public might not know any better, but companies ask for it when they just throw out numbers without sufficient explanation.

-- Anonymous, October 04, 1999

Lane,

And maybe Joey picked up a malicious hacker who took him way off track and ruined all his previous pretty plans and schedules. :-(

-- Anonymous, October 04, 1999

O.K., Lane, I'll grant you all of that. But just take a look at this from both a common-sense view and an IT view. (and I might not necessarily be redundant here!) If I have 20% of my budget left, AND if IT budgets are almost always under-budgeted for the job, (according to Mr. Yourdan), then doesn't it follow--from a common-sense approach here--that there is a lot more than 1% of the work left to do? (And yes, I know that these are some of the most grammatically awkward sentences I've ever written!!) :)

-- Anonymous, October 04, 1999

Oh, Ann. IT and common sense? You're way off base there. ;-)

Seriously, we just get numbers, with no clues to indicate how the money is being (or is going to be) spent. Maybe it's going for IV&V, maybe for planned purchases according to contingency plans, maybe for overtime for all those folks who'll have to be out in the field during The Rollover, maybe for an unbelievable mixture of anything and everything related to finishing up Y2K -- and for other things as well that they decide they can "hide" in the Y2K budget.

-- Anonymous, October 04, 1999