By request of some of you...

Here is a re-post of my Final Evaluation -- it is parts 2 and 3 combined.

Before we look at predictions, we need to review what we do know in various sectors of the oil industry. Those sectors we will look at include:

A. Oil well -- crude oil production (both domestic and foreign imports)

B. Oil transportation -- pipelines, ocean tankers and port docking facilities

C. Oil Refining -- refineries

D. Retail Distribution -- Your local gasoline station

NOTE: I do not have sufficient quantities of sources for me to confidently address the natural gas issues. There may or may not be cross-applications to consider but I will not deal with these. I will only deal with what I know with some level of strong likelihood or certainty. After reviewing these above areas, I will then present an evaluation of the data that suggests what outcomes may transpire and the percentage of likelihood.

BACKGROUND -- Reviewing the oil data

A. OIL WELLS and CRUDE OIL PRODUCTION:

Oil wells can be divided into 2 categories.

1. Large oil wells. These are loaded with automated computer systems that are composed of what are called "large-scale embedded systems." This automation reduces the need for human interaction and thus lowers labor costs. The automated systems are cost-effective because the well pumps enough oil to make it financially affordable. These types of wells account for nearly 80% of US domestic oil production and supplies about 36% of US needs.

2. Small oil wells. These are sometimes referred to as "stripper wells." These wells produce small amounts of oil. Stripper wells (generally) are found in older oil fields that were exploited long ago. It is not profitable for an operating company to install embedded systems onto these wells. These types of wells account for nearly 20% of US domestic oil production and supplies about 9% of US needs.

In regards to Y2K, "stripper wells" are and always have been compliant. They contain no automation. Large oil wells may have a compliance problem due to the large quantities of integrated computer chips installed within various embedded systems or Large-scale embedded systems (called LSES). The average large oil well will likely contain on average around 10,000 'chips' placed in approximately 50 to 100 Large Scale-embedded systems. Private industry assessments from preliminary testing indicated that 1 in 3 LSES is likely to experience a systems shutdown upon rollover. This does NOT apply to newer large wells that were developed within the last 2 years generally speaking.

Thus we have the potential risk of losing 36% or more than one-third of US supply needs if all of these wells fail. Virtually every large well runs this risk. Newer wells of the last 2-3 years make up a very small amount of the large oil wells and these are estimated to provide (at best estimate) less than 1% of that 36% figure. This leaves us with 35% being "at risk" for failures if not remediated.

The problems with remediation are multiple, according to my sources. The primary problem is testing for problems. Most oil companies do not know how many potential chip problems may be in a given well because each well was customized to unique situations present at that location. Many installations were not well documented or else those documents are now long gone. Many times wells are bought or swapped and over a 20 year period may have several different oil companies operating that well. Papers and documents tend to get lost in those situations.

Testing also necessitates that the wells be shut down. This is a serious problem in that in order to shut down a well the test results may be so severe that the well is permanently lost. In addition, testing a well could take too much time to do a realistic and effective job. Another problem is that when testing is performed and problems found, often times the chip cannot be easily accessed without destroying the sealed system in which it is placed. The system was not supposed to be repaired but rather replaced in toto should it ever fail before the well runs dry. In many cases, it was expected that the well would go down before the embedded systems. It is usually not the embedded system that goes out before the well does.

These systems are usually customized and are not something bought off of a shelf and placed directly into the operation. Most often they are built from scratch to specific order or modified from assembly line production. A customized system can take 3 to 4 months to build and install and test-out before placing online. Many of these types of systems have been referred to as inaccessible systems, because for all practical purposes it is inaccessible because they are sealed systems. To access it is to destroy the seal and because it is in a sealed system that cannot be resealed once it is tampered with. Also, within it, parts are coated for protection and cannot be identified because of the protective coating. It is a part of weatherproofing for the elements directly onsite. This doesn't necessarily apply to all the systems on a well, but it does to many. Some systems are also difficult to reach and may require disassembly of machinery to reach them, just like it may require disassembling part of your car's engine to reach certain computerized sensors that assist in controlling vital aspects of (say for instance) fuel injection. This is true also with an oil well.

Another problem in remediation is that some computer chips cannot be replace even if not in a sealed system because there are no longer replacement parts for that chip. Thus an entirely new system must be designed and constructed. Again we're looking at perhaps a 3 to 4-month lag time. Also, there are only so many folks qualified to do the designing and building of these systems. It is not likely replacing a transistor or IC chip in a VCR or TV set, nor like building a VCR or TV on an assembly line. There are no assembly lines for most of these systems being hung on oil wells.

So we have a whole series of problems involved in trying to identify and fix failures. We have a problem in lost revenue during a well's downtime. We have a threat of losing an oil well in testing. We have a problem in getting a chip or system replaced in a quick and timely manner.

We have a problem in terms of the rate of failures on the LSES on large oil wells. The rate of failures in LSES systems is 1 in 3 (average) on a well. There have been instances of 1 in 2 or half of the systems reporting failures. This means that anywhere from 15 to 50 systems may have severe problems that can shut down the well. Generally speaking there are at least 1 to 5 critical systems that can be expected to shut down that well due to date rollover problems. Many of the other system may fail but will not necessarily shut down production. The bottom line though, is that it is expected that at least one to five of the 15 to 50 systems are expected to have critical failure shutdowns. These are not very good odds.

B. Oil transportation -- Pipelines, Ocean tankers, and port docking facilities.

Pipelines are a primary means of transport for domestic crude oil. It provides a means of moving the crude oil from the wells to refineries where it is refined into gasoline, diesel fuel, kerosene, lubricants, etc. In ye olden days of yore, (about 20 years ago) pipelines began to undergo automation improvements designed to eliminate skilled laborers and high payroll costs. Automation was computerized utilizing the SCADA embedded systems. The SCADA systems in pipelines are not exactly the same as found in say the electric utilities. Pipeline SCADA systems are more complicated in applications and do indeed involve time-date issues at periodic checkpoints on the pipelines. Testing on an "end-to-end" basis could not be done without shutting down the entire system. This was not done. There was very little testing involved except for "type-testing." Based upon this limited testing, the industry has concluded that there is no problem.

Embedded systems experts in the oil business however, are privately not nearly so confident as the corporate PR news releases indicate. Frankly, from what I've been able to discover, it is truly an unknown status and a real wild card issue in assessing the likely impacts of rollover because there is so little that is truly known. In order to know, testing would have necessitated in disrupting the flow of oil in a major way that would have had severe economic impacts upon refining of crude oil products and thus affecting the entire economy by limiting refined supplies.

Ocean Tankers are also loaded with embedded systems in order to reduce labor costs by reducing the number of crewmembers needed to run the ship during voyage. Reports I've seen indicate as many as 150 LSES aboard and control a number of critical elements aboard the ship. Information regarding testing results is rather scarce and incomplete. Suffice it to say that reports I've read indicate that at least one or more systems are expected to experience problems if left unremediated. This has been a real concern for the US Coast Guard. The USGS has issued new rules mandating that ships that are not compliant will not be allowed to enter ports during or critical passageways during the rollover. I have not been able to gain enough data to assess the mathematical impact of this.

Here is a recent report on the US Coast Guard's position regarding this issue. From Federal Computer Week: http://www.fcw.com/pubs/fcw/1998/1207/fcw-newscoast-12-7-98.html Snippet:

" sAN DIEGO -- Year 2000 date-code problems in oil tankers could lead to an oil spill in a U.S. port or slow the flow of imported oil to U.S. communities, some of which have "only a three- or four-day supply of oil,'' the Coast Guard's chief information officer said last week."

Speaking here at the Navy Connecting Technology Conference, Coast Guard CIO Rear Adm. George Naccara said concerns about computer chips that control oil pumps on vessels and in on-shore storage facilities could force the Coast Guard to issue a directive telling tankers "not to offload cargo on Dec. 31, 1999." The Coast Guard can issue an order against oil offloading from tankers under its mandate to ensure marine safety in U.S. coastal waters and ports. The Coast Guard also is worried about how the Year 2000 problem will affect the marine transportation industry, especially the 7,000-plus foreign-flagged vessels, which make more than 80,000 visits a year to U.S. ports. Over the years, these ships have been automated with information systems that have allowed ship owners to cut costs by reducing the size of crews. Today, even the largest tankers can operate with only a crew of 10.

Naccara said the embedded-chip problems in some ships could come close to halting them dead in the water. "Some [vessels] may have problems crossing the ocean,'' which also could put a crimp in the nation's oil supply.

Here is a link to another Maritime website that spells out more details of problems with embedded system aboard ships:

http://www.ship2000.com/Ship2000.nsf/5a8ba7baee61c43c802565df004ad3ff/29a84b3de62b2a0e802566b300416c04?OpenDocument

Port Docking Facilities are also involved LSES equipment for loading and unloading oil. There are reports and indications that many facilities are not compliant and will not be compliant for rollover. Again, as with the tanker issue, I have no first-hand eyewitness sources indicating any information one way or the other on this. Based upon what I've read though, and the US Coast Guard information, there are apparently many port facilities around the world that are not ready. Reportedly, Venezuela and Saudi Arabia are having problems in gaining compliance, but this is subject to debate.

In the above-cited report in Federal Computer Week they also had this to say about on-shore problems

" The Year 2000 marine infrastructure problems extend beyond oil tankers, Naccara said, to items such as the giant cranes that offload containers from cargo ships. The Coast Guard also has started to develop contingency plans to mitigate problems with critical infrastructure systems, such as the electrical power grid.

Naccara said the Coast Guard has an ongoing "broad outreach effort'' to provide Year 2000 information to every segment of the marine community, including briefing the International Maritime Organization and distributing Year 2000 brochures to foreign-flagged tankers and recreational boats. Despite these efforts, Naccara said, "I have no doubt we have great exposure [in the marine transportation environment] due to the weakness of embedded chips."

End Snippet.

As you can see, there are some serious concerns raised by the USCG in this regard to oil tankers and shore loading/unloading facilities. One aspect that is scary that I didn't post was the comment from the International Maritime Organization of London spokeswoman who claimed that this issue of Y2K compliance was unknown to the safety division of that organization!! Hello????? See the article and also note the segments dealing with Chevron's exposure of its 35 tankers which were undergoing testing during December.

C. Refineries

Refineries in the US are primarily old facilities. Many date to the early part of the century and have been periodically remodeled many-times over. The larger refineries have large elements of LSES within them. This aspect of LSES in refineries is a closely guarded secret. It is one that the oil companies have not seen fit to publicly share. My insider sources tell me that most refineries with significant quantities of LSES are at severe risk. Specific refineries have been mentioned to me but sources have asked me not to reveal them because of possible repercussions to those sources who fear that with so few knowing the status the leak could/would easily be traced back to them.

Not all refineries are loaded down with LSES. Some have very few of these. Very small refineries do not have them in part because of the extensive costs and because they are owned by smaller independent oil companies. The Major Oil companies are the ones that have loaded down their refineries with embedded systems (LSES) in order to reduce expensive high-wage (union) jobs. Some refineries can be run with only a handful of men if heavily LSES'ed. There are two of the Major oil companies that reportedly have very limited exposure to the LSES problem and yet maintain larger to moderate sized refineries. These two oil companies are Marathon Oil Company and Ashland Oil Company. Both are subsidiaries of USX. (the old US Steel Corp.).

Most of the other majors BP/Amoco/Arco, Chevron, Texaco, Exxon-Mobil, Shell, Conoco, Enron, Gulf, have extensively equipped their refineries with embedded systems. However, again, not all refineries are equally loaded down to the same degree. Another compounding problem is that many refineries in the last year or so have been sold to independent or foreign companies. The major oil companies used to compose 70% of refining capacity. They are now down to 54% at last report. Independents now account for 23% of refining capacity. Mergers and acquisitions since these reports of a year ago probably have further altered those figures dramatically lower. Many of the independent companies have not had the deep financial pockets to perform remediation like the major oil companies. This may result in compounding the problems of Y2K remediation.

What do we know about Y2K remediation in refining? The oil industry is not providing details anymore about its progress. We can however refer back to this classic webpage for an interesting article on one Y2K remediation company's work with a major oil company. This article describes the remediation company's efforts and those results at the 18 refineries owned by the Client Company. We'll post below that some relevant quotes pertinent to our discussion.

http://www.prepare4y2k.com/embed1.htm

Begin article quotes:

"The results for the manufacturing systems were that 11% of the hardware, 14% of the software and 23% of the custom code was non-compliant. 2% of the suppliers were no longer in business and 12% of the suppliers would not supply an upgrade except through purchase of new models. "For the quality systems, 24% of the hardware, 29% of the software and 29% of the custom code was found to be non-compliant. All suppliers were still in business but 16% would not provide upgrades except through new models. The risk of failure was put at 60%.

"For the co-generation plant's results, 19% of the hardware, 36% of the software, and 24% of the custom code wasn't compliant. 13% of the products were no longer made. In this case, the energy management system tied into the local grid that was non-compliant and couldn't be upgraded. Some of the more interesting findings were that the catalytic cracker would fail and the refinery could no longer make gasoline. One of the more troublesome findings was that the analyzers would continue to work but would send erroneous data. The proprietary networks from the control systems to the analyzers would fail. The inventory and analysis would take 7 weeks and cost $122,000. The conversion for two units would take an estimated 15 weeks and cost $760,000.

"In Summary "In an overview of all these facilities, problems were found in the sensors & analyzers, lab equipment, programmable Control Systems, embedded systems, SCADA systems (Supervisory Control And Data Acquisition), distributed control systems, human-to-machine-interfaces (HMI's), networks, computers, third party applications, and Operating Systems. "Mr. Heerman also noted that they had found that some of the vendor compliance statements were incorrect. Some pieces of equipment the vendor had claimed to be compliant had failed. One piece of equipment successfully made the January 1, 2000 transition and was allowed to continue. Just over a month later, when checked again, the date on the equipment was January 34! This points out the need for continued testing, especially at the system level.

"The projected risk levels for failure of all the units of these companies was between 60% and 90% if the non-IT parts of the business were not found and fixed."

And another archived article confirms the premise:

http://www.cio.com/archive/090198_y2k_content.html

[snip] "Circuit boards aren't easily accessible," says Peter O'Farrell, a senior consultant with the Cutter Consortium, a research and consulting firm in Arlington, Mass. "You may have to completely dismantle the system to get at the circuit board. Then you can pull the chip off the board. But what if there's no new chip with the same footprint? Plus, you have all the static electricity issues, contamination issues and waterproofing issues. Messing around with boards is not much fun." Especially if they're located at the bottom of the North Sea. Ascent Logic Corp., a risk management technology firm in San Jose, Calif., has inspected Y2K projects at several oil companies with embedded chips on the sea floor that control wellheads. "Most of those chips will shut down on Jan. 1, 2000," says Ascent President and CEO Larry McArthur. "The question now is, Is there enough diving capacity and time remaining to get a significant portion of these replaced?" Be sure to order replacement parts and related services as soon as possible. "For every industry I've looked at, the order queues are filling up," says Roleigh Martin, a software engineer consultant in Edina, Minn. [end of quote]   

[snip]

"Many embedded systems are effectively "black boxes," purchased ready-made from a vendor without knowledge of the components or code that makes them work. "If we have a black box system controlling our HVAC system, no one within Amoco is likely to be intimately familiar with the contents of that box," says Hale. "As a result, you're very dependent on the supplier of that product for information that would inform you whether or not you have a Y2K problem."

[End of snippets]

My comments:

Now, the following is worth repeating again:

" One of the more troublesome findings was that the analyzers would continue to work but would send erroneous data. The proprietary networks from the control systems to the analyzers would fail. The inventory and analysis would take 7 weeks and cost $122,000. The conversion for two units would take an estimated 15 weeks and cost $760,000."

Did you catch that? "the analyzers would continue to work but send erroneous data." This tells me that some data will take a while to shut down systems. Others could be instantaneous. Note the length of time to repair this particular problem 15 weeks. That is nearly 4 months! Note also that (as I understand it) Problems like this could be hidden by instantaneous shutdowns that will mandate fixing before the other problems surface. So thatone item could shut things down instantlytake a long time to resolve and then have data corruption problems elsewhere surface perhaps 30 days after resumption of systemsthus shutting the systems down again for perhaps months.

So what are we saying? It is possible that refineries will have problems of perhaps of an immediate nature, and then again after repairs are made and systems resumedonly to see further data corruption (that had been halted) resumed without operator realizationonly to have further shutdowns days or weeks later.

In other wordschain reactions of problems in which we see a series of shutdowns in a chain that spreads out over time that could be hours, days or weeks apart. It could mean perhaps a half-year to a year of off again and on again shutdowns. In the case of refineries. You don't just restart a refinery by flipping a switch. It takes time to power up or rev-up a refinery, usually measured in days or weeks depending upon problems. This could be a real mess for refineries. Sometimes to clean up messes from a sudden shutdown can take perhaps weeks depending upon the severity of the problem and the extent of turnaround needs.

The stakes could easily result in extended losses of refined production levels across the oil refining industry.

D. Retail distribution -- your gasoline station

From what I've been able to uncover, this is one area where most of the industry is prepared and ready. Whether they are fully compliant remains to be seen. It is the one area where they faced the least amount of challenge, so I tend to think that most retail gas stations will be okayalthough perhaps their credit card systems may have problems that are perhaps related to difficulties with phone line access. Nevertheless, I am particularly optimistic in this particular area. This of course assumes that the power grid remains up and running.

PROJECTIONS FOR ROLLOVER:

CRUDE OIL: The following is based upon statistics and previously reported problems and failure ratios applied to general statistics. We'll look at each of the 4 areas that we previously examined only now we'll look at projections for each. From there we will draw overall conclusions that become our expectations for how the rollover might proceed for oil.

We are assuming that every other aspect of the national infrastructure experiences no problems. By this we mean that we are assuming that there is no problem with the national electricity grid and no regional or local power outages where refineries or ports are located. We assume the phones will be okay. We assume the water supplies to refineries are okay. We assume there are no problems with the banking or financial market sectors. We assume that all the vendors that supply support products to oil refineries, ports, tankers, and pipelines and oil wells are all okay. [that is a tall assumption, isn't it?] In other words, we are really removing from our equations any other potentials for trouble that are beyond an oil company's control. We want to try to measure just the main impact for the oil industry itself.

OIL WELLS and Crude Oil Production levels.

Given that 80% of USA crude production is subject to the Y2K embeddeds problem we realize the mathematical factors.

Small oil wellsno problem 20% of USA crude oil production or about 9% of total US needs will not be affected unless pipelines fail.

Large oil wells With 50 to 100 LSES for each well, and a 30% to 50% LSES fail rate per well we can expect 15 to 30 LSES to experience failures. This however does not necessarily mean that all the failures will shut the well down. In fact, we suspect that at least 1 system failure and probably upwards of 5 failures per well that can/will (?) cause shutdowns. This will not be for every large well as some wells newly developed had compliant equipment already. Still, in theory this should affect nearly 80% (give or take some) of domestic oil wells. Some wells may have manual overrides that can still work but others will not. The greatest problems will be with the large offshore wells. The Large Oil Well problem could affect as much as 36% of all oil supply needed daily by the U.S. (However, if Y2K affects other industries severely, there may be less need demanded).

For purposes of projections, Let us assume a mild scenario of only 1 in 10 oil wells suffers catastrophic shutdowns that can't be fixed in a few days. We know that 46% of our needs are met by US producers. Of that 46%  36% is from large oil wells. If 10% of supply is lost from those wells for indefinite or prolonged Periods we would then have a 3.6% supply drop to demand ratio. So we have a strong threat to the 36% of US oil supply going to refineries from US producers. A 10% reduction would mean a 3.6% drop in supply. This is roughly the equivalent of the 1973/74 Arab oil embargo but this is not the only threat. A 10% loss estimate is especially optimistic considering the mathematical odds but we can hope that parts are available in some cases, and workarounds available for other and the remainder of the 90% simply luck out. That would leave us with a 10% reduction. This, I say again, would be extremely optimistic given the mathematical odds based on the prior testing that gave us the failure ratios we are working with.

Foreign oil well production:

What goes for America, also works for oversees oil-producing nations. In foreign oil producing/exporting nations we find similar problems. This is apparently true for those cities that were highlighted in the latest US DoE "watch list" dated at the first of December. This list includes Venezuela, our number one source for imported oil. It also includes Columbia, Nigeria, Iran, Iraq, Russia and others. For the links and snippets see the old thread: http://greenspun.com/bboard/q-and-a-fetch-msg.tcl?msg_id=001zSh

See this link to the separate thread on Venezuela and the link to DoE on this nation. http://greenspun.com/bboard/q-and-a-fetch-msg.tcl?msg_id=001zHG

In addition, you will recall the other day's Bloomberg Financial News story citing UN officials reporting that Gulf Oil producing nations were not compliant. Later the story was downplayed and the spinmeisters came out of the woodwork to claim all is A-OK with oil. I've now had 3 seen three separate sources from oil insiders saying Saudi Arabia is in deep trouble and is not expected to make it even with their oil sector.

Folks, if the Persian gulf nations don't make their oil sector compliant and it goes Down, there is no need to continue this analysis. There's no need to even have this forum anymore, because if they go down, and hard as reports suggest, combined with all else you can kiss this nation goodbye. A 70% to 90% loss of oil would be the death knell of civilized society. So, we're going to take a more contrarian view and assume that it won't be that bad.

Let us optimistically assume, that only 10% of the oil production is affected in the Gulf. Let's project the potential problems on this basis. Please see the US DoE summary of Oil-importing nations and their percentage of supply to US needs. See the link: Latest monthly tallies:

http://www.api.org/faqs/

The link found below shows the US Senate figures on annual reports based upon 1998 final totals. (Remember that these are monthlies and fluctuate drastically from month to month.) Risks assessed by Gartner Group analysis supplied to the US Senate by request. http://www.iea.org/ieay2k/newlinks/imports.htm

We will also project that nations on the High level of probable disruptions will lose a third of their export quantities to the U.S based in part upon the mathematical factors plus the indications that those nations are less likely to provide sufficient infrastructure support. It may well be 100% and not 33%. We also then assumed that everyone else except the UK would lose 10%. (UK is not much of a factor anyway for US needs). What we then see here is a total impact of 22.8% of foreign crude oil supply lost to US. This translates out into an 11% of US needs not being met due to foreign interruptions. Combine this with a 3.6% domestic production loss and you have a 14.6% loss of crude oil. This would be far more severe than either of the previous oil-shock situations. Remember also, that I'm arbitrarily being optimistic in spite of the mathematical equations suggesting that all large wells will experience a shutdown failure. The prior testing models are suggesting that 1 to 5 of those systems that fail will be critical shutdown failures.

If only 10% stay down, that would be very optimistic. We note that parts problems are just one of many factors that could delay crude production stoppages. If you noted in the TAVA study, a real problem was lack of replacement parts because vendors were either no longer in business or no longer willing to make that part but rather required buying a whole new customized system. Well re-drilling is another aspect to consider when it has to replace the old well that lost all pressure and cannot be re-utililized. My point is that the odds are very much against just a 10% loss. But that is what we will work with except for high-risk nations known to have tremendous other infrastructure problems and severely behind on remediation.

What is the median expectation for well failures? I'd be more inclined to say a third to one half as being more realistic. If this were so then we'd be looking at a 30% to 46% loss of domestic US sources for US supply needs. Remember though that this reflects the deductions for the 9% factor supplied by non-embedded oil wells that must be subtracted.

So what do we have for an average projection range?

Most optimistic -- 14.6% loss of supply needs. Modest projection -- 30.0% loss of supply needs Median projection -- 46.0% loss of supply needs Maximum -- 91.0% loss of supply needs

When I review the statistics for the limited testing that was done and then apply that to the overall quantity statistics that we do have, I just don't see anything more optimistic than a 14% loss of supply, although I'd apply a plus or minus range of 5 percentage pointsfor a minimum range of 9% to 19%.

I also don't think it will reach maximum either. So, I'm still inclined to range it as between a 20% to 30% loss of supply based upon the statistics and insider reports of prior results.

B. Oil transportation -- pipelines, ocean tankers, port facilities

I simply don't have enough data to factor into projections for any problems that might be involved. I have noted the US Coast Guard commander's comments. We've noted that some port authorities are imposing restrictions. We've noted the problems with embedded systems within tankers and of course the loading and unloading systems in ports. Our projections then will assume these problems will only be short-lived. These assumptions may be too optimistic. I just have no statistics with which to crunch numbers for projections.

C. Oil Refining

This is the next big factor in projecting oil issues in Y2K. In doing so we should remember the earlier quotes from the Tava report on its experiences with its own major oil company.

Note the following sequence of failure???

"One of the more troublesome findings was that the analyzers would continue to work but would send erroneous data." (SO, IT WOULD CONTINUE TO WORKfor awhile, BUT)

"The proprietary networks from the control systems to the analyzers would fail. The inventory and analysis would take 7 weeks and cost $122,000. The conversion for two units would take an estimated 15 weeks and cost $760,000."

(NOTE: TAKE 15 WEEKS!!! That is almost 4 months!!!)

" One piece of equipment successfully made the January 1, 2000 transition and was allowed to continue. Just over a month later, when checked again, the date on the equipment was January 34! "

It took a month to show up!

" The projected risk levels for failure of all the units of these companies was between 60% and 90% if the non-IT parts of the business were not found and fixed."

Well those are pretty high numbers. We know that the US Department of Energy indicated in a 1998 year-end report that the US had 95 refineries operating last year. The report cited above refers to a major oil company that employed consultants from Tava to assist in remediation efforts. Prior remediation found a 60% to 90% risk of failure in embedded systems in these refineries if not fixed. We know, and the oil companies in their 10-Qs, as well as the NIST report tell us that they know they didn't/couldn't get them all. The US Dep't of Commerce's NIST report at the end of November tells us that testing was completely insufficient. So what kind of failure rates should we expect that would close down a refinery?

Now, I've come across information from an embedded systems oil engineer who served as the chief designer of embedded systems for a major oil company. He has over 20 years in the industry and until this past summer was involved in remediating those systems that he designed. He also assisted in helping other oil companies with their refinery embedded problems.

He has indicated that after reviewing the famous DD1 Light oil chat dialogue, he indicated that the projections made by DD are very "possible." He also said:

"No one (in this industry) is putting out accurate information any more, its impossible."

He goes on to say:

"When I was at ______ [Ed. Note: Oil co. name is deleted by this editor] site in ______, [Ed. Note: city location deleted for confidentiality] they said they were shutting down at the end of the year. There are a lot of other refineries doing same. "_______ [Ed note: Oil co. name again deleted] is having problems world wide. ARCO and EXXON are shutting down the two major international pipelines." "The least of my concerns are oil and chemical,  main concern is nuclear reactors. The same embedded systems used in the refineries are used at the nuke plants. The whole northern hemisphere is going to be exposed to radiation."

"Gulf is announcing they are shutting down overseas as non-compliant."

This fellow's experience with embeddeds in oil has led him to conclude that similar embedded problems in the nuclear power industry will lead to nuclear accidents and thus overshadowing problems within the oil industry. Indeed, if nuclear clouds were hanging over the whole Northern Hemisphere it would overshadow any percentage drop in oil supply or refined products. This is someone that I was hoping to make contact with and hopefully persuade him to come to this forum and share with us his knowledge. Unfortunately, I did not know how to contact him until a few days ago and the holiday time has made contact difficult although I have tried. I will keep trying to reach him and perhaps he will wish to post here.

The US Dep't of Energy in their 1998 summary says there are 95 active crude oil refineries operating as of the end of 1998 here in the U.S. What if 10% of those 95 US refineries go down and 10% of the foreign refineries go down? What is the impact? Substantial, to say the least. I will leave you to fill in the numbers, but what I've just related is a very, very optimistic projection. A single refinery shutdown in the past has sent tremors through the markets and triggered rising prices. Now, combine that with oil well production problems and you can see at least a pricing impact. This is based upon ignoring all other US refineries.

What if 20 refineries have problems? Especially if they are some of the largest capacity refineries. Keep in mind that a lot of the larger production plants have greater amounts of embedded systems and require only a handful of men to operate them. BUT, when a problem develops and a refinery full of embeddeds shuts down, there is a problem. It takes manpower to bring it back up. There are teams of specialists who go around doing nothing but starting back up refineries from maintenance turnarounds. Please note what another oil engineer had to say in commenting about the quotes from this engineer whom you just read about. This second engineer made the following comments about re-starting an embedded systems controlled refinery:

" Some crews specialize on such re-start-ups, so that shut downs are normally staggered to better utilize such specialists. An industry wide shutdown and fresh start-up is of itself an unprecedented hazard, because tiny misjudgments in start-up can be very costly, and destroy a plant. The hazards of having plants started-up without control by sufficiently experienced "start-uppers" might be almost as hazardous, and a global scale, as any embedded chip problems."

Now, perhaps you can see some other factors that are indirectly involved in this problem for the oil industry. Other problems can cascade indirectly from software code and embedded systems problems.

Do you think that perhaps my projection of only 4 or 5 plants shutting down is a bit too pessimistic? Or is it too optimistic? You decide. Frankly, I think I've been overly optimistic.

In summary, we don't know how many catastrophic events will occur on rollover that we will be able to see or hear about immediately. I suspect it could be several days. There will be delayed events that were fatally wounded but unknown to the company until perhaps a month later when it finally keels over dead surprising operators who had no clue. Expect many of these problems to take a long time to fix before production can resume. Yes, there may be some workarounds available but primarily in smaller less developed refineries.

Effects should begin to be felt later in the month at the pump. If it jumps in the first few days due to known problems, just figure it's only just begun to jump and that more problems are yet to come. Supplies would begin to shrink within days but the full magnitude may slowly develop at the pumps over a course of weeks. Probably mid-February if these systems follow the same fail ratios as during testing. How long will it last? I would suspect that it will be 6 months at the earliest to recover from the lengthy problems of system replacements and it is possible that it could be closer to 9 months or a year.

I suspect that we'll see at least 4 or 5 refineries out of production for at least a significant period of time (say 4 to 6 months). This would be the very minimum. That would be exceedingly optimistic. It is also premised on the notion that there are NO power outages or dirty power problems for any refineries. Dirty power could really screw up a refinery and quickly cause a shutdown. It also assumes that there will be NO phone outages, and NO water disruptions. Any shutdowns for any reason at a refinery loaded with embeddeds could mean a long wait for a special re-start team to comeIF there are several other refineries in the same boat. Remember a refinery that has gone cold can take a long time to get re-started.

Therefore, 4 to 5 refineries lost and out of production for an extended period of time is exceedingly optimistic. There is a good chance that 10 to 20 refineries could go down for various reasons, and not all might even be related to the embeddeds issues or power failures. It could be dirty power that knocks them out or perhaps phones or even complete loss of water for an extended period. My projection though is going to focus only on the embeddeds issue, not software code problems, not power loss, not dirty power issues, etc. We're just looking at it from the embeddeds standpoint. I think there's a good chance that we'll see 20 to 40 refineries go down with problems. Some will go down briefly, others for several days. These would not be the embedded refineries with only a handful of workers. These minimal staffed refineries will stay down much longer even if the embeddeds are fixed prontoIF many of them go down at once. Why? Not enough specialists around to restart.

I'm also not going to factor in possible explosions and fires that might take a refinery down for months or years. Still I think we stand a better than 50/50 chance for at least 9 or 10 refineries to go down for most of the winter and spring. Worst case would be all refineries going down for various reasons. I think that is highly unlikely without the grid going down. IF the grid goes down, the oil industry is TOAST for 6 months. PERIOD! Median case would be 30 to 40 going down and 20 to 25 staying down for 6 months or more.

It also assumes that pipelines will function normally and that there will be no problems with oil tankers or oil docking facilities.

Based upon a premise of no other infrastructure problems"

I think there is a 5% to 10% chance that no refineries will go down. I think there is a 5% to 10% chance that half the refineries will go down I think there is a 50% chance that 9 or 10 refineries will go down for 6 mo. I think there is a 20% chance that 20 to 30 refineries will go down for 6 mo.

I see a 95% chance for at least a 7% drop in oil supply for the USA I see a 70% chance for at least a 14% drop in oil supply for the USA I see a 50% chance for at least a 30% drop in oil supply for the USA I see a 60% chance for at least a 40% drop in oil supply for the USA

I see zero chances for NO oil. There will still be oil from stripper wells if the pipelines stay up.

I see problems developing fairly quickly but not publicly known for perhaps 2 weeks even if there are explosions and fires. I suspect the spin meisters will work overtime to blame any problems on any thing but Y2K due to liability factors. We will not likely know about oil and Y2K directly until later in the month, even if prices keep going up. There will be other excuses in order to keep the markets calm as long as possible.

I see 100% likelihood of LIES being spun by the government and big business starting on January 1, 2000  within 2 minutes after rollover. YOU CAN TAKE THIS ONE TO THE BANK.

-- RC (racambab@mailcity.com), December 29, 1999