BP Explosion 6

[0707]

From the BP provisional 57 of the 70 trays within the column were flooded in the moment of the first explosion.

The splitter column have been out of service during one month and steamed out to remove hydrocarbons from February 26 to 28. The steam is usually prone to tray valves because after steaming the valves stuck with scale and after start up the vapor pressure of the ascendant gases is not enough to unplug the valves and the column get flooded. The flooding of the column was responsible for the high peak pressure of 63 psig when the maximum relief pressure was 42. In my opinion the high pressure in the splitter was responsible for the hydrocarbons release followed by explosions in downward equipment.
When tray columns are shut down for long time, just after start up is important to mechanically knock the valves or flush the column with water to dissolve the scale around them. The steam degasification should be done only just after start up.

Luis Marques

 

[svanels]

I have read the interim the report (and with me many others), it is witten such way, that people understanding the technology clearly could make out what has been going on.
[ul]
[li]workers not following procedures (managers, ceo's are also workers)[/li]
[li]no supervision on critical moments[/li]
[li] paperwork that is signed (Management of Change) after parking trailers[/li]
[li]no crosscheck with outside instruments (column filling with no outlet and estimated time frame)[/li]
[li] people taking 100% readings for granted, would rise serious questions about the credibility of maintenance[/li]
[li]outdated technology (blowdown drum)[/li]
[li] tie-in instead of phase out of connections to blowdown drum[/li]
[/ul]
are all signs of management deficiencies, the traditional part of the iceberg that is underwater.
By the way if non-technical journalists are interpreting the report as:

BP PLC has backed off its earlier report that workers were primarily to blame for a March refinery explosion

you will get this mess, especially with a sue-happy public.

I saw in the report a summary of facts stated with evidence (no comments of spokesman this or union leader that).

 

[svanels]

The steam degasification should be done only just after start up.

I have been in only two turnarounds but I question this statement.
[ol]
[li]Shutdown[/li]
[li] Drain liquid in the tower[/li]
[li]Steam-out and drain condensate, this is done to get the column gas-free if you are going enter (confined space entry permit)[/li]
[li]Isolate!! Blank or Blinds list needed[/li]
[li] Inspection / Repairs[/li]
[li] Steam out again!!, you don't want oxygen in the system, hydrocarbons, oxygen and heat is an explosive mixture. That is why flare headers always have a slight positive pressure, if not lpg or fuel gas is spiked in to keep the air (oxygen) out [/li]
[li] pressure up with lpg or gas [/li]
[/ol]

The last step is start-up (introducing feed)

 

[0707]

Svanels

When a unit is planned for shut down, all degassing procedures should be followed. If columns are not supposed to be open for maintenance purposes, just after start up, is important to flush the column with water to dissolve the scale around valve trays to be sure that tray valves are free and able to tilt around tray holes to perform later its rectification job. If valve trays stay glued with scale, later on, the column will get flooded. The steam degasification to remove oxygen should be done only just after start up followed with nitrogen pressuring.
In Texas City the splitter was shut down on February 21, steamed out 26 to 28, drained on March 14 followed by nitrogen pressuring and depressurized on March 21. Cold feed was brought on March22/23.

Luis Marques

 

[svanels]

Our definition of start-up is when we are introducing the feed (hydrocarbons), I think that why the confusion. But why between steaming out and draining the condensate there are almost two weeks? Is it not better to pressure the unit up with nitrogen earlier?

Of course we don't know the time frame/problems they were experiencing, but from the report it was stated that 2 start-up procedures were used.
Just curious.

 

[unclesyd]

Some local fire department pictures and a little more information.

http://www.reliabilityweb.com/art05/bp_tx.htm

 

[jte]

unclesyd-

Good link. I received the slide show .pps a few days after the event. Pretty easy to note that it was not prepared on a BP/Amoco computer. Also when you take a closer look at the back of the firefighter turnout coats you'll note that not everybody belongs to the BP/Amoco refinery. It was a mutual aid situation. I'm wondering if anyone can substantiate the rumor that the person who shot the photos and/or released the slideshow is no longer employed...

jt

 

[jmw]

Just seen a report in one of the online magazines: "Pilot error"
BP is coming down hard on some operators and supervisors apparently.

JMW

http://www.ViscoAnalyser.com

 

[mrtangent]

All,

I have read the comments could you please answer some questions I had regarding the distillation column/ system.

I'm not familiary with the oil industry. So excuse me ignorance.

> Is Rafinate a trade name for mixture of heptane +.
> Whats the difference between light and heavy rafinate - BP are they both mixtures.
> It is common for oil industry columns not to have independent high level switches on distillation columns ? (I never design a column without one)
> on common vent stacks do you not normally have level alarms. (or even trips SIS2/3 .... /ESD 2,3)
> Why do they use fin fans not water cooled.

 

[svanels]

I am no chemical engineer, but IMO the heavy raffinate comes out the bottom, and the light comes out of the top. The light stuff is most likely a mixture of hydrocarbon gases.
It is also common to have high level switches that triggers alarms, emergency shutdowns etc.
Vent stacks like blowdown drums are dangerous, almost every where in the world they are substituted by flare systems.
Almost all hydrocarbon vapors are heavier then air, and will come down and acumulate in pits, sewers etc, if they are not flared (burned).
On the vent stacks there must be high level alarms, ESD's should be a difficult if different units are connected to the blowdown drum. Imagine 5 independent units connected, an upset in one of them could imply shutdown of all the others.

 

[svanels]

on common vent stacks do you not normally have level alarms. (or even trips SIS2/3 .... /ESD 2,3)

In case of a trip / ESD in a unit, the quantity of gases to be vented increases, since their normal stream is disrupted. Stagnant gases combined with heat will build up pressures that must be released to the stack/flare.

 

[IPGBen]

Hello,

if you use two independent level transmitters which are mounted separately on the column you can go without switches etc.
They must be used as independent signals etc. So each of the two must be able to trigger actions (e.g. alarms, ESD's etc.).

Grtz
Ben

 

[jsummerfield]

Actually lots of discussion is required to determine safety integrity levels for process units, systems and subsystems, etc. The generalizations for one facility cannot hold for another. A new facility might employ most of the following:
transmitters instead of process switches,
seperate systems for "basic regulatory process control" and safety shutdown,
different technologies for the measurement devices for the diffent systems to avoid common mode failures,
increasing levels of backup measurement where the safety integrety level for a process sub-system is elevated,
...

The unit at BP was not new. Typically, a process hazard analysis is required every five years to re-visit safety levels, monitor management of change, ...

The refinery codes and standards are typically not as stringent as required for offshore oil and gas production. Thus, some things that are done within some facilities are not requirements.

John

 

[mrtangent]

One question to savnels (Petroleum).

You state that stagnet gasses will need to be released to the flare stack does this mean that the column runs very cold ? or is there a reaction. Otherwise I dont understand why this is an issue. My understanding of distillation columns.

If you stop feed/heat then as long as the MAWP of the column is higher than the saturated vapour pressure of the lightest component at the base P then you dont have an issue. Or is this a Low MAWP column because its cold.

 

[svanels]

James we are dealing with hydrocarbons, which contains gases. Imagine a distilation train, consisting of heaters pumps, columns heat exchangers etc. Fired heaters are basically heat exchangers were crude passes through the tubes and an open flame on the "shell" side. We are talking about tube skin temperatures of near 600 C and product temperatures of 450 C or more.
If the heater feed pump trips, burners must trip also (interlock), the case is to get the feed running as quick as possible to prevent overheating of tubes and have a tube burst. As long as the product is running it is cooling also the pipes.

On the flare system various PSV's are coupled not only from the destilation tower. A trip results in temperature spikes in various points of the system, thus hydrocarbon gases are liberated which have to be removed by the flare system.
Whenever there is a trip the flames of the flare get higher, yellow with black smoke, indicating that an abnormal quantity of gases is flared.

Imagine you turn off the power switch of the whole plant (every operators nightmare). Theoretically you are not entering energy so things will cool off. The flare will be x-times higher then normal releasing "trapped energy"

The point is not only the distillation column, but the distillation train must be considered. There is no reaction involved in distillation (vacuum system) but when the temperatures run to high, coke will be formed in the pipe, which is a chemical reaction.

 

[owg]

The full (Interim)report is available online at ww.bp.com/texascityincident.

HAZOP at

http://www.curryhydrocarbons.ca

 

[owg]

Sorry that is

http://www.bp.com/texascityincident

HAZOP at

http://www.curryhydrocarbons.ca

 

[svanels]

Does anyone know about the final report?

 

[jte]

Don't know about BP's report plans, but the Chemical Safety Board is planning on holding a public meeting to discuss their preliminary report in September. I'd suspect that a final report would be issued a few months after that.

See second to last paragraph on page 4 of:

http://www.csb.gov/news_releases/docs/CSB BP Press Statement 6.28.05.pdf

Main incident page at:

http://www.csb.gov/index.cfm?folder=current_investigations&page=info&INV_ID=52

jt

 

[unclesyd]

According to the news BP had another excursion at the same site. The people near the site was just told they could come out of their house this morning.

I suppose there will be an addenda to the final report.