I've read the statements on this site by DD Reed and she presents a plausible argument about the oil industry not being ready. She states that Chevron and Tenneco admit that they wont be ready, largely due to embedded chips that are not Y2K compliant. This caused me some concern so I researched the subject as best I could.

First I checked the web sites of the major oil companies in the U.S. The results are listed below.

- Chevron  Chevron has a Y2K Readiness Statement on their web site but it is weak and unconvincing, particularly as to embedded systems. - Tenneco  I could not find a Y2K Readiness Statement on their web site. - Exxon  I could not find a Y2K Readiness Statement on their web site nor on the Exxon USA site. - Texaco  Has a Y2K Readiness Statement on their web site and it says what I want to hear. Spending $65MM. - BP Amoco  Has Y2K Readiness Statement on their web site and it doesnt say exactly what I wanted to hear, but they are spending over $300MM so at least they are working on it. - Shell  Has a Y2K Readiness Statement on their web site, but it is extremely weak. - Mobil - Has a Y2K Readiness Statement on their web site and it links to their 10Q SEC filling for details. Mobil has spent over $175MM and seems to be basically finished. I liked what Mobil said the best.

Second I talked to an independent geologist, now retired from the oil industry and he said that he had not heard of DD Reed, but that isnt too surprising. He was not aware of any serious Y2K problems in the oil industry but that isnt too surprising either. This conversation was somewhat reassuring to me, at least from the perspective that no news is good news.

Then I thought about one of DD Reed's main points that direct replacements for many non-Y2K compliant chips in use in the petroleum industry are not available. I dont know how many different chips are involved, but considering the dire consequences she details I would think that replacing those chips would command an emergency priority. If engineering and manufacturing designs are not available for those chips I feel sure that Motorola or Intel could reverse engineer and produce Y2K compliant chips within a couple of weeks. The information above shows that some oil companies are spending hundreds of millions of dollars on the problem. I would think that the effort to manufacture a few special chips would not be so great as to cause the demise of a few giant oil conglomerates. Its not reasonable.

However, having stated all that, I have decided to increase my food reserve from two months to six months. I have always been a big believer in failure avoidance and I think this is only prudent.

-- George Haas (gh451@hotbot.com), November 26, 1999

Answers

There have been numerous threads on this over on TB2000. Unfortunately, after reading it all, I'm still in the dark.

Yes, there are a lot of embedded chips. Everyone agrees on that.

Yes, if you had to replace them, you'd have to wait quite awhile for new versions to be manufactured. Probably simpler to replace a whole unit with a Y2K-compliant version. But parts shortages will mean long delays.

And yes, if you shut down a refinery or a pipeline, you may have major restart headaches. It could take months.

So yes, if we have lots of failures and no way to fix them other than replacing a unit, we're in trouble.

Some disagreement about whether systems are ever "down hole", or undersea. Some say this would be stupid, so it's never done. Others say that its common. No references to an equipment manufacturer site that describes such a unit. There are units with sensors in inaccessible locations, but it's never clear where the processors are.

Also not clear that any of these units have date logic, or that they could not be set back to prior dates if they fail. Some people think any chip with a real-time clock is going to fail. This can't be right -- if it has no way to set the clock, it's probably resetting to some base time when powered up, which has no relation to the current time. Think of an old DOS PC resetting to 1/1/80 and ticking forward from there (this is not unrelated to embedded systems -- many are single-chip DOS machines!)

Also not clear how much safety/operation critical stuff is computerized. A lot of infrastructure is really old, and built by people who (rightly) don't really trust software-based controls.

Very little insider information posted. Mostly it's friend-of-a-friend stuff.

My bottom line -- it's hard to believe there won't be failures somewhere. People can usually be counted on to make more than a few mistakes, especially when it comes to an unprecedented situation like Y2K. But catastrophic failures (50% drops in capacity?) Who knows?

I expect we'll see most actual failures on 1/1/00 though, with reports in the press by the end of January. Oil price speculators will do more thorough research than we can think of doing.

-- Michael Goodfellow (mgoodfel@best.com), November 27, 1999.

"If engineering and manufacturing designs are not available for those chips I feel sure that Motorola or Intel could reverse engineer and produce Y2K compliant chips within a couple of weeks."

I'll have to disagree with this point. I know nothing about the oil industry, but have a bit of experience in electronics manufacturing (mainly defense electronics). I've worked on emergency defense projects (one in particular during the Gulf war regarding ASW and mine detection), and some things can be turned around fast when the usual bureaucracy is instructed to lay off. However, doing a chip like that just ain't gonna happen. There are ways to "burn" chips on a very low volume basis, but you still have to understand the code and then duplicate the function. Some of that stuff gets so convoluted that it's more like a total redesign from scratch - assuming you even know what it's supposed to do in the first place. And that's a pretty big issue by itself.

Bottom line - it's in the same "wishful thinking" category as the "Bill Gates will rescue us" arguement.

-- Stephen McGehee (scm@adjutant.com), November 28, 1999.

Recent news has shown major suppliers (Mexico, Saudi Arabia, & Venezuela) stating that they will cover shortfalls caused by any glitches. BUT the supply chain is longer than the suppliers...and I'm not at all sure about a number of the major suppliers (Most of the Gulf states, Venezuela, Nigeria, just to name a few...). My feeling is that the statements are being made to assuage any panic...

I suspect that we'll see higher gasoline prices...and probably rationing.

-- Mad Monk (madmonk@hawaiian.net), November 29, 1999.

Without knowing the actual inner workings of these "embedded chips" I'll have to just throw out a concept. What if a program were written to work like a virus and go into any exposed systems and modify/trick/ or otherwise cause the chip to just go on as before.

-- R. Larry Heyer (vonlar@nextdim.com), November 30, 1999.

Unfortunately, it doesn't work that way. Even the crudest amateur virus still has to know EXACTLY what it is looking for, what format it is in, etc., in order to work. Unless you've done programming, I can see how this would look like a workable solution, so it's certainly not a "dumb idea". Software (and firmware) just doesn't work that way. In fact, that's part of the whole Y2K problem - the system is expecting a very precise piece of data and there is no tolerance for deviation from that expected data format. Computers are very good at doing things the same way every time, and doing it extremely fast. They're also really dumb. Again, nice idea, but just not workable.

-- Stephen McGehee (scm@adjutant.com), November 30, 1999.