On Mon, 7 Jun 2004 17:03:49 +0100, "Keith Willshaw"
wrote:


"Peter Skelton" wrote in message
.. .
On Mon, 7 Jun 2004 15:33:05 +0100, "Keith Willshaw"
wrote:

For the plant it might be, and there might be casualties there,
but, because the gas is over its ignition temperature, it can't
BLEVE. You get a fire burning the material in the pipe. There
wouldn't be effect beyond the fence. If the plant systems
functioned properly, the outage might be less than two weeks.

BTW, how would you go about breaking this line? A buldozer isn't
going to get there. These lines are fairly robust and plant's
just in case defenses against leaks are considerable.


Oh come on Peter. There are LPG lines all over the dammed
place on any refinery and a major leak is bloody hard
to contain.

The regulations must be drastically different over there.

Go and look at the report on what happened at
Flixborough

I have, in detail, often, with access to a lot that isn't
generally available.

(I can think of much worse scenarios, but not ones started by a
bulldozer that begins outside the fence. They start with operator
or maintenance error compounded by control room error.)

In the case of the Flixborough accident in the UK a pressure
vessel was bypassed by the maintenance dept using pipes and bellows
units.
Unfortunately the bypass was not properly anchored
and a slug of liquid caused the bypass to tear loose.

Flixborough happened in 1974. At that time, I was employed by
DuPont at Maitland ON, a plant that has a very large Cyane
oxidation unit so we had passing interest in the event. IIRC,
they were using a temporary bypass that had been constucted
without engineering assistance. A slug of process fluid, caused
by a process upset, tore a bellows that was improperly installed.


Gee I just said that

No, you blamed the accident on failure to anchor a bypass.

There was no automatic shut-off upstream. The plant lacked modern
process controllers[3] and was, even by standards of the day, not
centrally controlled.


Quite so , not that it would have helped much

It would have ended the fire within fifteen minutes.


The explosion was 15 tons equivalent of the BLEVE [1] type, the
fire lasted days becuase about 10% of the plant inventory had to
be allowed to burn out [2]. There was minimal effect past the
fence.


Wrong. Even though the explosion occurred on a rural site
53 members of the public received major injuries and
hundreds more sustained minor injuries. The plant was
destroyed as were several others on the same site and
close to two thousand houses, shops, and factories
were damaged with some 3000 residents being left homeless


No part of the plant met modern standards.

There are plenty of 1970's pterochem plants
still out there and the best control system in the
world doesnt help when you dump 50 tons of
Cyclohexane into the environment.

There aren't may fifties plants out there and there aren't any at
all that will dump fifty tons of cyane from a pipe rupture.

The causes of the
event were internal to the plant. The process affected was
obsolete and hazardous at the time and recognized as such.


A bulldozer tearing open a line would have
had the same effect.

How do you get the bulldozer to the line? Then how do you get the
line to dump much more than its contents? And who still oxidizes
cyane outside a collum?

The situation you describe is nothing like this. In your case
vapour burns as soon as it finds an oxidizer, mixing is not
possible. Shut-offs would function automatically and limit the
amount of fuel. There will be no big bang, although there would
be one hell of a whoosh.


You are assuming no coincident or consequential damage occurs, this
is a POOR assumption. What structures are being weakened
by that flame and what happens when they fail.

No an awfull lot. That's what the controlls are about.

BTW, I'm assuming the builldozer doesn't get far into the plant.
It's not all that easy to do here.

It is such risks that are rarely analysed and often
provide the nasty shock when an incident occurs

One of the worst industrial Bleve's happened on a
french plant where a small fire started at a faulty valve.
Trouble is the flame impinged on a LPG storage sphere

BANG


You've still not dealt with the basic question. Which is whether
there was a chemical plant near the incident that was so grossly
mis-constructed and mis-managed as to be vulnerable to such an
attack.

The furnace scenario you chose shows little understanding of
explosions or chemical plants. The plant you chose is ludicrously
different from existing types.

The CPI is not immune to accident. There have been many, there
will be more but this is a low-probablility scenario. In the case
at question, calling in an air strike because of the possibility
that the bulldozer might enter a chemical plant and do mischief,
I'll stick with what they decided to do.

Peter Skelton