BY RAYMOND E. HARING
On March 10, 2003, at 3:37 p.m., a propane transport tanker carrying 9,200 gallons of product exited Interstate 664 North to U.S. Route 17 North in Suffolk, Virginia. As the tanker rounded the off ramp, the driver lost control of the vehicle. The vehicle’s tank portion rolled over the guardrail while the tractor portion remained on the road side of the guardrail. The tractor and tank remained connected and slid on the driver’s side of the tank for approximately 176 feet. The vehicle came to rest with the tractor lying on the driver’s side with the tank attached to the tractor also lying on the driver’s side and down the off-ramp embankment.
The Suffolk (VA) Fire Department responded to the scene and immediately secured the area around the vehicle. An engine company crew was sent in to rescue the driver and to determine if there were any leaks. No leaks were detected. The driver had exited the cab, sustaining only minor injuries.
 (1) The 30-inch dent in the tank where the weld attached the front of the tank to the barrel, as seen during the walkaround. (Photos by R.C. Powell.) |
null
 (2) The tank after it was uprighted after the conclusion of the incident. For safety reasons, the contents were removed before the vehicle was uprighted. |
null
Suffolk Fire Department Battalion Chief William Edwards arrived at the incident and took control as the incident commander. He called the Virginia Emergency Operations Center (VEOC) in Richmond, advising of the situation and requesting to speak with a Virginia Department of Emergency Management (VDEM) Hazardous Materi-als Officer (HMO).
The VEOC paged me at 4 p.m., advising me of the situation and that Chief Edwards was requesting a telephone call. I called Chief Edwards and was told that the tanker was on its side and not leaking but that there was a faint odor of propane in the area immediately around the tanker. I told Edwards I would respond to the incident and assist with assessing the damage to the tank and determining the safest method for righting the tank and tractor.
I arrived at the incident at 4:22 p.m. and spoke with Chief Edwards and Captain Richard Schmack, the incident safety officer. Also on scene was Trooper R.C. Powell, a motor carrier officer from the Virginia State Police. We discussed whether the tanker could be lifted and placed back on its wheels safely or if the propane needed to be offloaded prior to lifting. I suggested Schmack and I make an entry to check for leaks and complete a damage assessment of the tank so that an informed decision about how and when to move the tank could be made. Edwards concurred and authorized an entry to perform a tank assessment.
At 5:10 p.m., wearing full turnout gear and SCBA, Schmack and I entered the hot zone to assess the damage to the tank and evaluate the feasibility of lifting the tank with the propane onboard. An engine company was in place with cover lines, should a fire occur. As we did an initial walkaround, the tank appeared to be in good shape with only minor road scrapes or “road burn.” It looked as if the tank could be picked up fully loaded and that the incident could be concluded in just a few hours.
However, when I lay down on the ground to inspect the underside of the tank that was not visible from the walkaround, I discovered a 30-inch-long dent just forward of the weld that attached the front head of the tank to the barrel. The dent was very sharp and was located in a critical area of the tank, just forward of the weld in the heat-affected zone (see photos 1, 2). Given the radius and location of the dent, the potential risks were significantly upgraded. We exited the hot zone and reported our findings to Edwards. I explained the seriousness of the dent and that if the tank were lifted with the propane onboard, there was an increased risk that the tank might fail. I added that given the position of the tank in relation to the valves and internal piping, it would not be possible to pump propane in the liquid phase from the tank; that the safest course of action would be to burn off the propane using a flare stand. Edwards agreed that the product should be removed before righting the tank and that if flaring the product would safely accomplish the removal, the operation should be set up.
During the next several hours, fire departments and propane distributors throughout southeastern Virginia were called to locate a flaring device. Although two area fire departments did have such a device, the supply hose was just 1/2-inch in diameter and too small for the job. A flare system was finally located around 8:30 p.m. at AmeriGas Inc, in Richmond. Nick Arvin, regional engineering and safety manager with AmeriGas, contacted me and said he had the proper equipment and would dispatch a truck and a technician from the Richmond facility to assist in setting up of the flare.
 (3) A four-burner flare. |
null
 (4) |
null
While waiting for the flaring equipment to arrive, I told Edwards that firefighting crews and a sufficient water supply would be needed during the entire operation. No hydrants were located within a mile of the incident, so a portable tank and several tankers were brought in to ensure an adequate water supply.
I requested that the Portsmouth Fire Department Haz Mat Team, one of four fire departments comprising the Southside Tidewater Regional Haz Mat Team, respond to the incident. On arrival, the team was tasked with bonding and grounding the tanker to eliminate any uncontrolled static discharge or spark.
The AmeriGas technician arrived with the flaring equipment at approximately 11:00 p.m. (see photos 3, 4). The Suffolk firefighters, the AmeriGas technician, and I worked together to set up the flaring equipment. The flare device was placed 120 feet uphill and upwind of the tanker. A supply hose was connected to the pump discharge valve on the propane tanker; all connections were double checked to ensure they were tight.
At 12:15 a.m., I informed Edwards that all was ready for the flaring operation to begin and that his firefighters needed to be in place should something go wrong when the flare was ignited. At 12:20 a.m., the flare was ignited. Propane flowed from the tanker to the flare and burned, producing a large flame approximately four feet in diameter and 30 feet tall. This continued for some time; then the size of the flame began to decrease. The haz mat team, using a standoff infrared thermometer, determined that the propane inside the tank was steadily cooling, thereby reducing the amount of vapor being discharged and burned in the flare. This was expected. Everyone knew that the cooling process would extend the burn time significantly.
At 6 a.m., I told Edwards that the burn rate could be modestly increased by applying water to the outside of the tank, thereby introducing additional heat into the tank to accelerate the production of vapor. I also presented the idea of pumping water into the tank through the tank’s spray-fill connection. Adding water to the tank would cause the propane to float on the water, which would lift the liquefied propane up to a level that would cover the pump discharge valve. Bringing liquefied propane above the discharge valve would allow the propane to be burned in the liquid phase and greatly decrease the time needed to burn off the product. The chief approved the plan and asked me to coordinate with Captain Schmack to implement it.
Schmack and I stretched a 21/2-inch hoseline from an engine to the propane tanker. A hose from the propane tanker was connected to the spray-fill valve. At that point, the first significant difficulty of the incident was identified. We needed an adapter to connect the 21/2-inch fire hose thread to two-inch pipe thread. A pipe shop at Newport News Shipbuilding fabricated an adapter and delivered it to the incident site at 8 a.m. The hoseline with a 21/2-inch smooth bore nozzle was then connected to the hoseline going to the spray fill valve (see photos 5, 6). The nozzle was used as a valve to control the flow of water into the tank. One large tow truck was placed next to the tank, and two slings were placed around the tank to prevent the tank from sliding down the embankment as water was added and propane burned off. Propane weighs 4.3 pounds per gallon; water weighs 8.3 pounds per gallon. As water was added, I did not want the increased weight to pull the tanker down the embankment.
At 8:40 a.m., all personnel on-scene were briefed on the operation and what to do should an emergency occur.
I advised Edwards at 9 a.m. that all was ready to introduce water into the tank. The temperature of the product was measured at 34°F. Propane is such that the pressure inside the tank is approximately two times the temperature of liquid; the pump operator was instructed to charge the water fill hoseline to 70 psi. Chief Edwards ordered the operation to begin. Water began to be pumped into the tanker. With no way to determine the exact level of the propane, a change in the appearance of the flame was used as the indicator to determine when liquid propane was flowing to the flare. A change in the flame’s appearance from clean-burning to somewhat sooty and a dramatic increase in heat output indicated that liquid propane was being burned. Water was added for approximately 10 minutes, at which time a large increase in flame size was seen and soot began coming off the flame (see photos 7, 8). At that time, the water flowing into the tank was stopped and the flame watched. At 9:30 a.m., the flame again became lazy, and more water was added. Again, the heat and fire increased as liquid propane flowed from the tank’s pump discharge valve. This process continued throughout the morning. The tank’s temperature was continuously monitored and showed a slow but steady rise during the morning hours.
Between noon and 2 p.m., progress slowed significantly. Although it was thought that water was entering the tank, it wasn’t because the propane had warmed during the morning to 50°F while the pressure on the water supply line remained at 70 psi, which was not enough pressure to overcome the approximately 100 psi in the tank; consequently, no water entered. Shortly after 2 p.m., the water supply line pressure was raised to 100 psi. The flare immediately responded, and the burning of the liquid propane was again underway. The operation then continued uninterrupted into the evening.
During the entire time that the liquid was being burned off, a hoseline was kept flowing on the area around the flare to keep the concrete road surface and surrounding area cool. This required a large supply of water. An 8,000-gallon water tanker from the Carrollton Volunteer Fire Department was brought in, along with two tankers from the Suffolk Fire Department.
With no way to determine the level of the water/propane interface, it was anticipated that at some time during the evening the water level in the tank would reach the tank pump discharge valve. At that time, water would begin spraying from the flare and the valve supplying water to the tank, and the valve coming off the tanker pump would be closed and the fire allowed to extinguish itself. This occurred at 10:20 p.m.
 (5) The connection between the apparatus hoseline and the pipe leading to the tanker required a special adapter (arrow). |
null
null
After the residual propane in the flare burned out, the flare was cooled with a hoseline. Next, the tank pump discharge valve was opened, and water was bled from the tanker to a level below the pump discharge valve. The valve was then closed. At that point, the flare was again ignited, and the remaining propane—all of which was vapor—was burned off. The flare self-extinguished at 11:40 p.m.
At that time, no propane remained in the tank but had been replaced with 9,200 gallons of water. The flaring apparatus and water fill line were disconnected from the tanker. The tow-truck operators arrived and used two heavy-duty tow trucks to lift the tractor and tank and place them back on the off ramp. The pump discharge valve was opened, and the water was allowed to flow out of the tank.
 (7) The propane burner in operation. |
null
 (8) |
null
The incident was terminated at 3:15 a.m. on March 12, 2003. No injuries or deaths resulted from the operation, which was determined to be an overwhelming success. At the beginning of the operation, it was anticipated that it would take between two and three days to burn off the propane; but using the technique outlined here, the removal of the product took only 23 hours.
In the book Propane Emergencies, by Michael S. Hildebrand and Gregory G. Noll, it is noted that flaring vapor using a two-inch hose would take 54 hours to burn off 11,000 gallons of propane. By using the method described above, adding water to the tank and burning the liquid propane, such an operation’s time can be reduced by at least half.
- Know your local resources and how to access them. This incident was successfully concluded through the efforts of fire, law enforcement, haz-mat, and industry personnel.
- Become familiar with the design, construction, and safety features of high-pressure cargo tank trucks, such as those used to transport LPG and anhydrous ammonia.
- Because of their inherent design and construction, high-pressure cargo tank trucks built to the MC-330/331 specification can take a significant amount of stress and damage and maintain their integrity. Responders must know where and how to access product or container specialists when this expertise is required.
- The Propane Emergencies curriculum is an excellent training resource developed by the propane industry for the fire service and emergency response community. Fur-ther information on how to get copies of the Propane Emergencies textbook and related training materials can be found at www. propanesafety.com.
- The equipment used at this incident was the bare minimum. Equipment such as a sight glass to monitor water flow and a remote shutoff device are strongly recommended, if available.
- In my experience, MC-331 tankers tend to dent or buckle at the same location as the tank in this incident dented and should always be closely examined after an accident.
LESSONS LEARNED
RAYMOND E. HARING has been a hazardous materials officer with the Virginia Department of Emergency Management since 1998, responsible for supervising the operations of two regional haz mat teams that cover 14 jurisdictions in southeastern Virginia. Previously, he had worked with the Virginia Beach (VA) Fire Department Haz Mat Team for 11 years.
PROPANE EMERGENCIES: MAKING A DIFFERENCE
BY GREGORY G. NOLL
In 1999, the National Propane Gas Association (NPGA), in cooperation with the Propane Education and Research Council (PERC), developed Propane Emergencies, a comprehensive training program for the fire service. Its primary goal is to improve firefighter safety in responding to propane emergencies. A partnership between fire service/hazardous materials respon-ders and propane industry product and container specialists developed the program. A team of fire service instructors and responders subsequently provided additional technical support and final review.
Now in its second edition, the Propane Emergencies curriculum consists of three elements: the Propane Emergencies textbook, a trainer’s Facilitator’s Guide, and a dedicated Web site (www.propanesafety.com).
Textbook. Written by Mike Hildebrand and myself and funded by propane industry assessments paid to the PERC, the propane industry’s financial commitment supported the creation of a 220-page textbook covering the following topics:
- propane standards, codes, and regulations;
- physical properties and characteristics of propane;
- design and construction features of bulk and nonbulk propane containers;
- design and construction features of bulk transportation containers;
- bulk plants and bulk storage tanks;
- general emergency response procedures; and
- tactical response guidelines for propane emergencies, including 20 typical emergency scenarios involving propane.
More than 100,000 copies of the textbook have been distributed at no cost to the fire service since its initial release in 1999.
Facilitator’s Guide. A comprehensive Facilitator’s Guide for trainers and instructors supports the Propane Emergencies program curriculum. Devel-oped by Michael Callan, the package includes comprehensive lesson plans for course deliveries in eight-, 16-, and 24-hour formats. The curriculum includes a CD-ROM with lesson plans and interactive training scenarios, overheads, slides, and a full-scale animated PowerPointT presentation.
In addition, the training package is supported by a 50-minute video produced by the Emergency Film Group of Plymouth, Massachusetts. The videotape is divided into two major segments: Segment 1 focuses on the properties and characteristics of propane; Segment 2 focuses on tactical considerations using the Eight Step Processq as the framework. The video will be available in a DVD-format in the future.
Web site. The program also has a dedicated and comprehensive Web site, www.propanesafety.com, that provides a program overview, instructional tips, background information on how to make training props, up-to-date changes to lesson plans, and downloadable graphics support for the instructor.
For more information on the Propane Emergencies program, including textbook and curriculum ordering information, fire departments and other emergency response organizations should access the Propane Emergencies Web site at www. propanesafety.com.
GREGORY G. NOLL, C.S.P., is a senior partner with Hil-debrand and Noll Associates, Inc., an emergency planning and response consulting business. He is a certified safety professional (CSP) and co-author of the Propane Emergencies textbook and six other hazardous materials textbooks, including Hazardous Materials: Managing the Incident. Noll is the assistant chief of Lancaster County Haz Mat Response Team and a member of the Fire Engineering Editorial Advisory Board and the FDIC Educational Advisory Committee.
DAMAGE ASSESSMENT OF PRESSURIZED CONTAINERS
BY GREGORY G. NOLL, C.S.P., AND MICHAEL S. HILDEBRAND, C.S.P.
Responders can be confronted with a variety of pressurized containers, including bulk and nonbulk cylinders, cargo tank trucks (MC-330/331), and railroad tank cars (e.g., DOT-105, 112, and 114 specification tank cars). Common products they may contain include propane, butane, and anhydrous ammonia. Much of the available literature on damage assessment of pressurized containers is based on testing and experience with railroad tank cars. Therefore, it should be recognized that tank car design and construction standards are substantially greater than those for cylinders and cargo tank trucks and provide a greater margin of safety.
More than 30 years of accident experience has shown that many pressurized containers have sustained extensive damage in accidents without releasing their contents. This is especially the case with MC-331 cargo tank truck and railroad tank cars equipped with head shields and shelf-couplers. However, improper offloading or uprighting techniques can increase the potential for mechanical failure of a container even if the outer shell shows no significant structural damage. One often-cited example of this concern was the Waverly, Tennessee, train derailment in February 1978 (Fire Engineering, May 1978, p. 39) in which an LPG tank car had a delayed catastrophic release some 42 hours after the initial derailment.
CAUSES OF CONTAINER FAILURE
The violent rupture of pressurized containers can be triggered by two related conditions: (1) the thinning of the tank metal as a result of scores, gouges, and thermal stress; and (2) the presence of a crack in the tank metal associated with dents. Factors affecting the severity of container damage include the ductility of the container metal (i.e., the ability of the metal to bend or stretch without cracking) and the internal pressure causing stress on the container metal.
Emergency responders should have a basic understanding of ductility. Ductility is the relative ability of a metal to bend or stretch without cracking. Ductile metals have a fine-grain structure and tend to bend but not crack. Brittle materials have a coarse-grain structure and tend to crack rather than bend. When a ductile container cracks, the crack tends to be small; in contrast, a crack in a brittle steel container tends to run linearly and cause the container to fail.
The ductility of a container metal is affected by the following:
1. Shell specification. This is particularly important in evaluating the integrity of damaged railroad tank cars, since different types of steel with different characteristics have been used over time.
2. Steel temperature. The higher the temperature of the shell at the time of damage, the more ductile (less brittle) the steel will be and the less risk there is for failure. For many liquefied gases that are loaded cold, it will take time for the shell temperature to rise to ambient levels.
3. Cold work. This refers to the deformation of steel when it is bent at ambient temperatures or that results from an impact or static load. Cold work reduces the ductility of the metal.
4. Heat-affected zone. This is the area in the undisturbed tank metal next to the actual weld material. This area is less ductile than the weld or the steel plate because of the effect of the heat of the welding process. This zone is most vulnerable to damage, as cracks are likely to start there.
DAMAGE ASSESSMENT
In conducting damage assessment of pressurized containers, evaluate the following factors:
 |
Road burn or street burn occurs when a pressurized container overturns and slides some distance along a cement or asphalt road. It is a long dent that is inherently flat, and container damage is typically superficial unless accompanied by sharp dents or gouges in combination with a weld area. [Photo by Lancaster County (PA) Haz Mat Response Team.]
- Container type. Identify the type of container (e.g., DOT specification number), construction material, and internal pressure. Methods for determining the internal pressure include pressure gauges and temperature gauges with pressure conversion charts.
- Stress factors. Determine the type of stress applied to the container (e.g., thermal, mechanical, or a combination of stressors).
- Stability. Evaluate the stability of the container. Use caution when inspecting an unstable container, since it may move or shift during the inspection process. It may be necessary first to stabilize the container with blocks, cribbing, air bags, or other means.
- Surface damage. Examine the container’s surface, paying attention to the types of damage and the radius (i.e., sharpness) of all dents. Railroad personnel will often use a tank car dent gauge as a “go/no go” device for comparing the radius of curvature of a tank car dent to accepted standards and to determine the severity of damage. However, note that the tank car dent gauge is NOT applicable for assessing damage to cargo tank trucks.
Experience shows that the most dangerous situations will include the following:
- Sharply curved dents or abrupt dents in the cylindrical shell section that are parallel to the long axis of the container. The container should be offloaded without moving it when a dent is considered critical.
- Dents accompanied with scores and gouges.
- Scores and gouges across a container’s weld seam. If the score crosses a welded seam and removed no more than the weld reinforcement (i.e., that part of the heal that sticks above the base metal), the stress is considered noncritical. However, if the score removes enough of the base metal at the welded seam, then the stress is considered critical.
- Cracks in the base metal of a tank or cracks in conjunction with a dent, score, or gouge. If the crack is leaking, the situation is very serious. These situations, as well as others, justify offloading the container as soon as it is safely possible.
Although public safety personnel can use these guidelines to make general decisions about uprighting vs. offloading or flaring, technical assistance from cargo tank specialists should be obtained and become part of the hazard and risk evaluation process.
Damage assessment is, at best, a qualitative assessment process that requires input from product and container specialists who are experienced in evaluating structural damage and the integrity of pressurized containers (e.g., stable vs. nonstable, go/no go, etc.).
Endnote
1. Hildebrand, Michael S. and Gregory G. Noll. Propane Emergencies (Second Edition). Chester, MD: National Propane Gas Association (2002).
GREGORY G. NOLL, C.S.P., and MICHAEL S. HILDEBRAND, C.S.P. are senior partners with Hildebrand and Noll Associates, Inc., an emergency planning and response consulting business. Both are certified safety professionals (CSP) and are co-authors of the Propane Emergencies textbook1, on which this sidebar is based, and six other hazardous materials textbooks, including Hazardous Materials: Managing the Incident. Noll is the assistant chief of the Lancaster County Haz Mat Response Team and a member of the Fire Engineering Editorial Advisory Board and the FDIC Educational Advisory Committee.
