Monday, June 30, 2008

Barton Solvents Fire Report

Last week the Chemical Safety Board issued their final report on the storage tank fire at the Barton Solvents’ fire in Valley Center, KS. As had long been expected the cause of the fire was determined to be a static discharge inside of the storage tank. What is important from a chemical safety point of view is that the Board determined the root cause of the discharge and disclosed a significant, apparently widespread, potential safety issue. This safety issue could be easily turned into the operational equivalent of a vehicle-borne improvised-explosive to attack chemical facilities.

As a Process Chemist with 12 years manufacturing experience I am certainly interested in the safety issues that the Board has pointed out. In all fairness to the companies that I worked for, these issues are not new; we took them into account in the Process Hazard Analyses’ and Process Safety Reviews that we conducted. We evaluated every flammable organic liquid that we handled for static accumulation hazards; when data was lacking we assumed it accumulated static charge. This is one of the reasons that we nitrogen blanketed the head space over flammable liquids as a matter of course.

Chemical Safety Issue as a Facility Security Issue

As I noted in a blog shortly after this incident (see: "Security aspects of Barton Solvents fire in Valley Center, KS") incidents like this point to a number of potential security risks. Little did I know then that the Chemical Safety Board, in their case study of this incident, would point to a new class of chemical storage targets. According to the CSB Barton Solvents Case Study (page 5):

  • "VM&P naphthas, however, and other flammable liquids (e.g., many NFPA Class IB Flammables), may form ignitable vapor-air mixtures inside tanks at normal handling temperatures."

This is unusual. Most flammable liquids have such a high vapor-pressure that the headspace in storage tanks is so saturated with flammable gas molecules that there is not enough oxygen available to support combustion. The atmosphere in the tank is above the flammability limit for the chemical.

What this means is that storage tanks containing these chemicals identified by the CSB may have a bomb in the head space just wanting a detonator to set them off. What is very scary from a security point of view is that there are a number of very common industrial chemicals that form these ignitable vapor-air mixtures. According to the study these chemicals (IVAM chemicals, a new term of mine) include (but are certainly not limited to):

  • Cyclohexane
  • n-Heptane
  • Benzene
  • Toluene
  • n-Hexane
  • Xylene
  • Ethyl Benzene
  • Styrene

Chemicals Not Regulated Under CFATS

None of these eight IVAM chemicals are on the list of Chemicals of Interest (COI, found in Appendix A, 6 CFR part 27) that DHS uses to evaluate a facility to see if it is at high-risk of terrorist attack. That means that facilities that store significant amount of chemicals in this group may not be required to complete a Top Screen unless they have at least one other chemical on site that appears on that list and is above the Screening Threshold Quantity (STQ) for that COI.

Furthermore, even if a facility did meet the requirements for being designated as a high-risk facility for some COI on site, the facility would not need to consider the presence of any of these chemicals on site. This would allow the facility to overlook the possibility of a terrorist using a simple attack (incendiary bullet shot into the headspace of the storage tank) on an IVAM chemical storage tank as a means to catastrophically release a toxic COI in a nearby storage tank.

IVAM Chemical as a Terror Weapon

The Chemical Safety Board, as part of their final report, released a safety video (follow the link on the report page) showing the lessons learned from the Barton Solvents’ Fire. This video should be viewed by every security manager and facility manager for a facility that has storage tanks containing IVAM chemicals.

Watching the animation of the naphtha storage tank being launched by the initial explosion is impressive enough. Watching nearby storage tank lids being blown off by the expansion of gasses being produced by the boiling liquid contents is truly scary. Fortunately, in this incident the other storage tanks only contained flammable and combustible liquids. They did not contain any toxic release COI. The problems of fighting this fire would have been greatly complicated by a plume of toxic gasses.

Anyone who watched this drama play-out over a couple days last summer will remember the impressive camera shots of the black smoke cloud rising over Valley Center, KS. The flat, open terrain made it look even more impressive with the flames flashing through the base of the smoke. If there had been a message of responsibility released by a terrorist group shortly after the start of the fire, the national coverage would have been much more intense. Local news shows would cut in with local experts discussing the vulnerability (real and imagined) of local chemical facilities.

If a nearby storage tank had contained a toxic-release COI, the news coverage would have gotten insane. The relatively organized evacuation of the nearby neighborhood would quickly have turned to panic as some people began to be affected by the chemical cloud, both physically and psychologically. The news coverage would have switched with music-video rapidity from long shots of the ‘toxic’ cloud, to local news crews filming and interviewing panicky residents ‘fleeing for their lives’, to contract talking-heads in New York and Washington exaggerating the effects of the ‘dangerous industrial chemicals’.

Mode of Attack

All of this could be set off with a relatively easy rifle shot using military grade, or even civilian available, tracer rounds. The walls of non-pressurized storage tanks are typically just sheet metal that could be penetrated by a pistol round. A high-speed rifle round would fly right through and potentially penetrate multiple tanks.

The tanks at Barton Solvents were exposed and relatively close to a public road. They were certainly close enough to the road to be engaged by a rocket propelled grenade like an RPG-7, the second most common terrorist weapon after the AK-47. A gunner standing in a sun roof of a car or in the back of a pick-up truck stopped momentarily in front of the facility would have been high enough to clear the fence while targeting the top couple of feet of the storage tank.

Defending Against the Attack

While the attack would relatively simple to execute there are a wide range of defenses that could be employed to prevent the attack.

A high privacy-fence could be erected at any fence line along a public road; this would increase the difficulty in targeting the tank. A tall sheet metal fence around the tank farm would achieve the same effect and probably require fewer linear feet of fence. Neither fence would have to provide ballistic protection; they would be designed to interfere with targeting not to stop bullets.

Ballistic protection could be provided for the IVAM by adding Kevlar® jackets to the storage tanks. The facility could also use bolt on ceramic or steel armor to provide ballistic protection. The engineering would be fairly straightforward, but the price would be higher than fencing discussed earlier.

A very easy solution would be to keep oxygen out of the headspace of the storage tank; without oxygen there could be no fire. Keeping a nitrogen, or some other non-flammable gas, blanket on the headspace of the tank would prevent the explosive ignition of the IVAM chemical. This could not be used for monomers like styrene; they frequently require the presence of oxygen to stabilize the chemical. For those chemicals a collapsible bladder could be placed in the headspace of the tank to exclude the monomer molecules from the headspace instead of excluding oxygen.

DHS Needs to Re-look at Flammable Release COI

As I have said on a number of other occasions, I believe that DHS has been overly selective in the flammable liquids that they have included in their flammable release COI. This newly identified ‘class’ of flammable chemicals could easily be added to Appendix A, using the CSB report as justification.

Lacking that, DHS should require facilities currently beginning the SVA submission process to report the IVAM chemicals they have in storage tanks within some pre-defined proximity to any toxic-release COI storage tank. That way this ‘release-trigger’ for the toxic COI can be identified.

No comments:

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