Cross-Post from CYS2k

From: Anonymous 9:20 PM

Subject: Another plant goes BOOM!

Wednesday June 23 7:05 PM ET

Two Dead In Phillips Petroleum Chemical Plant Blast

HOUSTON (Reuters) - Two people were killed and three injured in an explosion and fire at a Phillips Petroleum Co. (NYSE:P - news) chemical plant in the Houston area Wednesday.

Phillips spokesman Vernon Johnson said it was not yet clear what had caused the explosion, which occurred in a part of the plant that produces a material called K-Resin.

Production at the K-Resin unit, with a capacity of 375 million pounds per year, has been halted for an undetermined period.

K-Resin is a clear, shatter-resistant styrene-butadiene copolymer that is used to make products such as medical device components, toys, food packaging and clothes hangers.

-- (fake@out.com), June 23, 1999

Answers

http://dailynews.yahoo.com/headlines/ts/story.html?s=v/nm/19990623/ts/phillips_fire_1.html

-- (fake@out.com), June 23, 1999.

interesting...there was also a small fire here in Southern California at a Mobil facility in Torrance that was knocked down quickly. No idea yet what caused the problem, at least not in the news reports...

Mike =================================================================

-- Michael Taylor (mtdesign3@aol.com), June 23, 1999.

Do these two 'accidents' have any tie-in to a Y2K related failure? I am not being critical, just looking for the link. Stuff like this happens everyday somewhere in the world-has been for a long time.

-- Barry (barryjaynes@usenvitech.com), June 23, 1999.

No further information available yet on the Houston explosion. Maybe someone in the Houston area has heard more on their local news programs??

-- (fake@out.com), June 23, 1999.

Hmm...kind of reminds me of yesterday's headline:

Air Force Y2K Test Goes Swimmingly: Pilots Swimming in Ocean After Two F-15s Crash During Test

-- a (a@a.a), June 23, 1999.

Barry, who knows? But there have been a whole lot of explosions and fires lately. Much more than I realized before. Maybe I'm just more sensitive to the reporting. Maybe getting news from all over the globe regarding these blasts makes me "feel" that it is happening more often. Or, maybe some of these are directly related to testing, failure, glitch?

Mike ===============================================================

-- Michael Taylor (mtdesign3@aol.com), June 24, 1999.

A timely article:

Could Y2K cause an accidental chemical or nuclear spill?

-- Jack (jsprat@eld.net), June 24, 1999.

link doesn't work for me.

-- (fake@out.com), June 24, 1999.

Lets try that again!

Link

-- Jack (jsprat@eld.net), June 24, 1999.

There are chemical spills, explosions, fires, etc. even without Y2K. Expect that the problems will be even worse, after...

-- Mad Monk (madmonk@hawaiian.net), June 24, 1999.

Is there anywhere where all serious chemical plant failures are (publicly) recorded? If so, it should be easy to establish whether there is a rapidly rising trend. (A general web search is no use: a rise may indicate either increased quanitity or merely increased reporting)

I still suspect there isn't an increase, it's just that this community is now sensitized to the ones that do happen. Chemical plants never have been the safest of places - my father had plenty of stories to tell about near-disasters back in the 1960s!

-- Nigel Arnot (nra@maxwell.ph.kcl.ac.uk), June 24, 1999.

Attorney Leon Kappelman discusses chemical plant explosions with Ed Yardeni. Timely! He raises the issue that EPA and OSHA, regulatory bodies of the chemical industry, presented legislation to address y2k related dangers at the 177,000 chemical plants in the U.S. They concluded that 85 million Americans who live work or play w/i 5 miles of a chemical plant are at risk. The Senate totally ignored the recommended legislation and in fact passed opposing legislation.

You need Real Audio to listen:

http://www.y2kactionday.com/audiofiles/yardeni/t200/kappelman.ram

Read about the issues at:

http://www.chemsafety.gov/

-- OR (orwelliator@biosys.net), June 24, 1999.

Don't remember the specific thread, but we looked at the relative frequency of oil and chemical plant fires -

after oct 1998, the rate of fires and explosions went up 6 times: fires or explosions were six times more frequent than before that date (worldwide). These latest two continue that trend.

-- Robert A. Cook, PE (Kennesaw, GA) (cook.r@csaatl.com), June 24, 1999.

By the way - thinking of "linked effects" and medicine - for example -

< K-Resin is a clear, shatter-resistant styrene-butadiene copolymer that is used to make products such as medical device components, toys, food packaging and clothes hangers. >>

Oooops - they need a new supplier - don't they? Got a spare chemical plant in your back yard? --- Flint - yes, I'm exaggerating, there really are two or three suppliers - worldwide - for these sort of raw goods. I just hope next year they don't all fail at once.

-- Robert A. Cook, PE (Kennesaw, GA) (cook.r@csaatl.com), June 24, 1999.

Thank you Robert, once again, the voice of reason. Please read between the lines, folks. We've been basically told, "you have nothing to fear from embedded systems...keep your money in the bank...we're working very, very hard and really, really fast...everyone please, just sit down and shut up". It just gives me such a warm, fuzzy feeling.

-- Will continue (farming@home.com), June 24, 1999.

I agree with Nigel... absent any historical tracking data, we're really just engaging in idle speculation. Does anyone have actual proof that the number of "accidents" has risen above normal levels?

-- M.C. Hicks (mhicks@greenwich.com), June 24, 1999.

Well M.C., you may refer to Robert's post above (never known him to lie) and then hit the archives. OR...look up 'utilities' in the archives and find the 'Kansas City Power' thread, OR you may continue to rely on the media for absolutely nothing but 'accidental' reports of the numerous events, lacking their ability to connect the dots. If someone is able to provide a hot-link to a study done...I'd be thrilled. I consider these 'hints' to be a deciding factor of what is around the corner.

-- Will continue (farming@home.com), June 24, 1999.

Osha tracks them. But their data only cover 1989 through 1997. Can't find the recent data.

-- fake (fake@out.com), June 24, 1999.

From CNN:

Blast, blaze hits Phillips plastics plant in Texas

June 24, 1999 Web posted at: 10:16 AM EDT (1416 GMT)

HOUSTON (AP) -- An explosion and chemical-fed blaze erupted inside a plastics plant, killing two workers and injuring four others.

Wednesday's blast rocked the section of the Phillips Petroleum Co. that makes clear plastic. The dead men, a maintenance mechanic and a pipefitter, were employed by San Antonio-based H.B. Zachry Co., company spokeswoman Cathy Green said.

Green said their names were withheld until officials could notify the workers' relatives. The four injured men were treated and released from hospitals.

The cause of the blast has not been determined, said Earl Borths, the plant's industrial relations director. Federal inspectors were expected to investigate Thursday.

The plant was the site of a deadly explosion in October 1989 that killed 23 and injured 130.

H.B. Zachry is under contract with Phillips to provide industrial maintenance and services at the plant, located in the Houston suburb of Pasadena, Green said.

Zachry officials did not know about the two deaths until midafternoon, when a final count of employees was made after the blast, said Green.

Calls to Phillips in Houston were forwarded to the company's Bartlesville, Oklahoma, headquarters. Officials there didn't return a call from The Associated Press.

-- Roland (nottelling@nowhere.com), June 24, 1999.

Here's a link that summarizes refinery explosions.

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

In summary:

10/19/96 Illinois, Blue Island, Clark Oil

11/11/96 California, Los Angeles, Texaco, hydrotreater

1/22/97California, Martinez, Tosco

6/22/97 Texas, Houston, Shell

8/25/98 California, Martinez, Tosco

10/16/98 Pennsylvania, Trainer, Tosco, jet fuel tank

11/25/98Washington, Anacortes, Equilon PugetSound, "coker"

1/13/99 Arkansas, Smackover, Cross Oil, Naptha Valve

2/24/99 California, Martinez, Tosco

3/25/99 California, Richmond, Chevron

5/7/99 Texas, Pasadena, Lyondell-Citgo, coker

5/13/99 Corpus Cristi, Coastal

-TECH32-

-- TECH32 (TECH32@NOMAIL.COM), June 24, 1999.

Oh! This doesn't include the gasoline pipeline explosion a few weeks ago, nor does it include the explosions elsewhere in the world. The worlds largest gasoline refinery (or so the article said) in South Korea went up a few months back as did a biggie in India.

Got diesel?

-TECH32-

-- TECH32 (TECH32@NOMAIL.COM), June 24, 1999.

So, in the US alone, there were 4 reported fires/explosions between Oct 1996 and July 1998, or 4/21 months=0.190 fire per month.

Between Oct 1998 and June 1999 (9 months) there were 10 explosions (including three in June: the Bellingham, WA gasoline pipeline rupture, the Torrance CA fire, and the Houston TX explosion and fire.) This is 10/9 months = 1.111 fire per month.

5.83 times the previous rate. I'd say that's statistically relevent.

-- Robert A. Cook, PE (Kennesaw, GA) (cook.r@csaatl.com), June 24, 1999.

Will -- Don't get me wrong, I am VERY suspicious about the cause of these incidents. Absent proof to the contrary, I tend to assume worst case (that Y2K is a factor).

TECH32 -- nice post! The link doesn't work, though (not for me, anyhow). Updating it with recent events, it does seem as if the pace is picking up. Was there any summary information regarding the "official cause" for all these events?

-- M.C. Hicks (mhicks@greenwich.com), June 24, 1999.

Robert -- Yep... seems mighty relevant to me. Also seems like it might even be the type of "hard data" analysis that gets mainstream media to take notice.

-- M.C. Hicks (mhicks@greenwich.com), June 24, 1999.

And then there's the utility explosions/fires: Ford Rouge etc

-- (fake@out.com), June 24, 1999.

M.C. Hicks,

Try cutting and pasting the url. The link got screwed up when it wrapped.

-TECH32-

-- TECH32 (TECH32@NOMAIL.COM), June 24, 1999.

Don't forget the Ford plant! (right?) Perhaps some of US should 'connect the dots' and inform the media of their ignorance or sloppy investigative 'slump'. As a matter of fact...since the government is well barricaded and the availability of 'the vote' may be in question, we are left with 'the media' as our only source of battle. There are many choices other than Clinton News Network (CNN) or the other big money waste-of-times. When a person approaches the Mayor of their town about Y2K, and his response is, "What's that?"......it just may be time to consider other avenues (that is a true-life account, experienced by me and hubby last September...go figure) Is there NO reason at all for hot-blooded people like myself, to be a bit fed up? Please...let's put some heads together. There must be something we can do, besides arguing with dead-heads and bashing the clueless.

-- Will continue (farming@home.com), June 24, 1999.

Last night, I distributed the postings on this thread to my e-mail distribution list. Got back a very interesting reply, which I've gotten permission to post. The respondent asked to remain anonymous....

* * * [Note: ABENDS = "abnormal ends"]

Michael,

The whole topic of tracking faults/errors/ABENDS/snafus is very difficult.

In 1987 I worked in a well known financial services firm. As part of my indoctrination process I was assigned to attend the daily 8am production debriefing. Every day about 15 people would crowd into a tiny room suitable for maybe 10 (actually helped to get us in & out in a hurry) to review what had happened in that night's production.

They also maintained a quite detailed on-line log of incidents. Whenever a job-step hiccuped it would be logged in.

Management was increasingly concerned that as stock market volumes rose, the rate of production incidents followed right along.

I then proceed to analyze six months of incidents by putting them into a spreadsheet. (This was an exercise, since our 40 person department didn't even have a single PC at the time, so I had to go over to the training department which was beginning to have PCs!)

Lo & behold patterns emerged! There were something like 6 errors that consistently accounted for about 2/3rds of the production incidents. And all of these error types were quite mundane... a file running out of disk space, bad JCL, bad user input. The rate of programs ABENDing from poor testing was almost non-existent.

At this point, thought I then ran into the political buzzsaw. From the preliminary results I had, it was pretty clear that with some more detailed monitoring of precisely where & why these common errors were happening, it should be pretty easy to buy large amounts of time... unfortunately this analysis ran completely counter to the motivations of the department head. He was more interested in using the rising number of production problems as justification to hire more people.

There were other details about these error logs that showed how difficult it was to really know what was going on. Even though there was a detailed log on an incident, there was no record of how bad a problem it'd been. Was incident X a 5 minute outage or two hours? That had to be dug out manually. The error logs tracked that an incident had happened, but not how much time/effort was required to resolve it.

The basic situation comes down to the fact that the development staff (where I've always been) was in fact far removed from operations. There was very little feedback between the two organizations.

I had this precisely driven home recently when speaking with a Y2K project manager friend. He's been winding down their three year Y2K effort. Two thirds of their 90 mainframe systems needed to be Y2Kok by 1/1/99 & made it. For a period of several months in the beginning of the year he told me there had only been 4 production incidents traced to a Y2K bungle (resulting in outages of 4 hours, two of 8 hours each & one for 20 hours). When I asked him how that compared to the normal rate of outages in their shop, it took him a bit of time to dig up what the normal baseline is.

Which goes to your point of... does OSHA (or anyone) actually have accurate numbers of the NORMAL rate of refineries/chemical plants blowing up. I assume such complex plants are having problems all the time. Just like men probably continue to die in coal mines on a steady basis, but it's just not news anymore.

Ultimate gross example is deaths on the highway... we continue to kill something like 40,000 people annually & don't even bat an eye.

[Name withheld, by request]

-- M.C. Hicks (mhicks@greenwich.com), June 25, 1999.

So, in the US alone, there were 4 reported fires/explosions between Oct 1996 and July 1998, or 4/21 months=0.190 fire per month.

Between Oct 1998 and June 1999 (9 months) there were 10 explosions (including three in June: the Bellingham, WA gasoline pipeline rupture, the Torrance CA fire, and the Houston TX explosion and fire.) This is 10/9 months = 1.111 fire per month.

5.83 times the previous rate. I'd say that's statistically relevent.

-- Robert A. Cook, PE (Kennesaw, GA) (cook.r@csaatl.com), June 24, 1999.

And you call yourself an engineer? The only way that it is statistically "relevant" is if your list is comprehensive and if the time periods used are comparable and all inclusive as well. Where did this list come from? Just the ones that people could think of and post? Big mystery that there would be more recent ones listed since people seem to forget things after a while. I know of a big explosion at a Georgia-Pacific resins plant in Columbus, OH in 1997 that didn't make your list and there are probably dozens more in both periods that are missing.

-- Do You See (howstupid@youlook.com), June 25, 1999.

M.C. Hicks: Workplace fatality stats available from OSHA at http://www.osha.gov/oshstats/work.html

*WorkPlace Injury and Illness Statistics

Incidence Rates

Incidence rates(1) of nonfatal occupational injuries and illnesses by industry and selected case types, 1997 1996| 1995| 1994| 1993| 1992| 1991| 1990| 1989 Occupational Injury & Illness Data for 1997 Occupational Injury & Illness Incidence Rates per 100 Full-Time Workers 1973-97 Chart of Workplace Injury and Illness Incidence Rates, 1973-1997 Search the BLS Incidence Rates 1986-Present

Fatalities

Fatal Occupational Injuries by Industry and Event or Exposure, 1996 (PDF/33K) Fatal Occupational Injuries by Event or Exposure and Major Private Industry, 1996 (PDF/15K) Fatal Workplace Injuries, 1974-1991 Fatal Workplace Injuries, 1992-1997 Census of Fatal Occupational Injuries The Evolution of Occupational Fatality Statistics in the United States (PDF 528KB)

Injury/Illness and Fatality Characteristics

Keyword Search of Available BLS Injury/Illness and Fatality Data, and Publications Lost-Worktime Injuries: Characteristics And Resulting Time Away From Work, 1995 (PDF 84KB)

BLS

Bureau Of Labor Statistics (BLS) Safety and Health Statistics Home Page Industry Injury and Illness Data - 1996 Case and Demographic Characteristics for Workplace Injuries and Illnesses Involving Days Away From Work- 1996

-- fake (fake@out.com), June 26, 1999.