Sunday, August 16, 2015

Chemical Facility Fire in Texas

Friday afternoon there was a fire at a chemical facility in Conroe, TX. The facility was a supplier of drilling chemicals for the oil field. Interesting news accounts here, here, here, here and here.

Background

There are a wide variety of chemicals used by the crude oil drilling industry. Many of the products used contain flammable solvents; including toluene, xylene, methanol and acetic acid. The first two are not soluble in water and typically float on water. The second two are water soluble and in fairly low concentrations make water flammable. Applying water to fires involving any of these chemicals has a tendency to spread the fire, not put it out.

Oil field drilling chemicals are typically shipped to the field in containers; 5-gal pails, 55-gal drums, and 250-gal plastic totebins. From the outside of the facility we cannot tell if this was strictly a warehouse of if blending operations happened in the facility. We can only see three relatively small storage tanks outside, but there may have been additional tankage inside of the building.

This facility was not specifically designed as a chemical warehouse or chemical manufacturing facility. According to Google Street Views as late as February 2013 it housed an insulation and fireplace supply company. As such it was originally designed to have a sprinkler system. From the progress of the fire (described below) that water based system may have been functioning on Friday.

The Figure below is a diagram that I drew of the facility based upon Google Maps. It is consistent with the photos shown in the various news stories about the fire. Bldg 1 was principally a warehouse with five loading docks facing the parking lot. The north end of Bldg 2 was also a warehouse facility with three loading docks. The south end of Bldg 2 was the facility office. There were roll-up doors from both buildings facing into the space between them with a ramp leading down to the parking lot level.



The Incident

From the news reports and accompanying photographs we can piece together much of what happened at this facility. A full investigation is underway and the initial cause of the fire is unknown.

At about 4:00 pm CDT the company closed up business for the weekend. The last employee left and the gates were locked. Apparently about 45 minutes later the fire started. Pictures (here) seem to indicate that the fire started in Bldg 1. By the time this picture was taken it is clear that there had been a release of one or more flammable liquids in the building and it had started to flow out of the building since you can see flames on the concrete parking lot.

There are a number of reports of explosions associated with this fire; with at least one being described as ‘large’. With fires in chemical warehouses it is very common to have containers ‘explode’. The heat of the fire causes the liquid to boil inside the container. The expanding gasses (even water vapor) in the container cause the container to catastrophically fail creating a small explosion. If the container contained a flammable or combustible liquid the expanding gas cloud would ignite providing a larger explosion. The relative sizes of the two explosions would depend on the volume of the container and the amount of solvent in the container.

A later picture shows the parking lot fully engaged in flames. Again this is a sure sign that there has been a major spill of a flammable liquid. Fortunately the parking lot was designed to keep any liquid on site and flowed into the drainage basin located on the south end of the facility. Aerial photos (here, here, and here) show that drainage basin on fire. From those photos it looks like the major fire in the parking lot was out by the time that the fire department arrived on the scene.

Pictures from the aftermath of the fire (here, here, and here) would indicate that Bldg 1 was a total loss, there is severe damage to the warehouse portion of Bldg 2 and there does not appear to be any significant off-site damage or runoff.

Probable Course of Fire

This was an unusual chemical warehouse fire. Fires that start in these facilities after hours typically involve electrical systems or non-chemical debris on site. For there to have been a significant chemical release early in the fire without a catastrophic explosion (the building was intact in early pictures) is very unusual. To get fire flowing into the parking lot there had to be a large amount of flammable liquid released; more than a drum or totebin’s worth.

I suspect that there was at least one storage tank inside the building containing a product with a fairly low concentration of flammable solvent (so that there wasn’t a large explosive vapor cloud). Somehow there was a failure of that tank that allowed the contents to start to drain onto the warehouse floor. At some point (either before or after that leak was initiated) a fire started igniting that liquid on the floor and the sprinkler system tripped applying water to the fire. The water from the sprinkler system spread the fire throughout the warehouse and out the door leading to the parking lot ramp.

The fire would have spread to wooden pallets holding drums or totebins of other combustible or flammable liquids in the warehouse. As those liquids started to heat there would have been a number of drums or totebins that would fail and some of those would have resulted in small fuel-air explosions as the volatile solvent vapors ignited.

The remaining liquid in those containers would have also been washed into the parking lot by the sprinkler system, contributing to the pool fire there.

At that point the warehouse would be fully involved and nothing would stop it from burning until all of the fuel (including chemicals that are normally rated as not being combustible) was consumed.

The large volume of fire in the parking lot was almost certainly caused by the failure of the small storage tank at the north end of Bldg 1. The smaller pool fire would have spread to near that tank. It looks like it was a plastic tank so that the fire caused the bottom of the tank to soften and release the contents. From the size of the resulting fire ball, I would assume that this was a solvent tank and resulted in a large fuel-air explosion described in the various news reports.

Commentary

Fortunately, this facility was designed with a system to catch rainwater and return that water to the aquafer via the drainage basin. This system was not specifically designed for catching chemical run-off from the facility as it is clearly visible in the pictures taken before the current occupant moved in. This allowed all of the burning runoff from the fire to be contained on site. The size of this fire would have been significantly larger if that system had not been in place. It will also make the clean-up of the aftermath of this much easier to accomplish.

It appears that the large pool fire in the parking lot was substantially over by the time that the fire department arrived. If fire trucks or personnel had been anywhere near that parking lot when it was fully engaged they would have been destroyed. Firefighters approaching a chemical facility fire really need to be aware of drainage patterns at the facility before they approach too closely.

Finally, facility owners and fire departments need to look at alternate routes of access to these types of facilities. In this case the wind was light and out of the east, blowing the smoke away from the only access to the facility. If the wind had been out of the north or worse yet, northwest, there would have been no way for firefighters to safely approach the fire.

Counter-Terrorism Notes

Now I don’t know exactly what products this facility contained, but I would suspect that this was not a facility that would have been required to report to DHS under the Chemical Facility Anti-Terrorism Standards (CFATS). That means that it is not really fair to discuss site security measures for the facility. The facility did have a perimeter fence and locked gates, much the same as you would see at most industrial facilities across this country.

Having said that, this fire would have been fairly easy to have started as part of a terrorist attack. Approaching through the woods behind the facility, a lone attacker with some small explosive devices could have started a nearly identical fire by putting those small charge on a number of different totebins containing flammable chemicals scattered around the warehouse.


A small fairly isolated facility like this would not be a typical target form Islamic militants or radical militia members, but an environmental wacko (no, not an environmental activist, but a real fringe nut case) would find a company associated with supplying the crude oil drilling industry a prime target.

No comments:

 
/* Use this with templates/template-twocol.html */