KANSAS GRAIN DUST EXPLOSION
BY BILLY JACK WENZEL
Editor`s note: Additional information for this article was supplied by Chief Gary Curmode and District Chief Greg Thompson, both of the Sedgwick County (KS) Fire Department.
At 0918 hours on June 8, 1998, a grain dust explosion rocked a grain elevator in a rural area of Sedgwick County, Kansas, two miles outside of Wichita. The incident reinforced numerous lessons, among them the extreme effects of dust explosions and hazards they pose; the creative nature of rescue firefighting and the critical roles training and experience play in collapse environments; and the importance of a strong unified command system for handling dangerous large-scale incidents.
The elevator was of steel-reinforced concrete construction. Its walls were 15 to 24 inches thick. The elevator in all its aspects was 2,700 feet long and 100 feet wide, making it the world`s largest grain elevator under a single head house. It contained a complex system of conveyors, bins, and storage areas for the offloading, transfer, and storage of large amounts of grain. Each of the many bins held approximately 70,000 bushels of grain when completely full. The head house, a 225-foot-high structure, housed the technology from which workers controlled the flow and placement of grain, but areas above the bins called “galleries” also were used for product control.
The explosion occurred on a Thursday morning and is believed to have originated in the tunnel area. The blast traveled through two tunnels and upward through empty bins to the gallery level, effectively wrapping itself around the head house, leaving certain floors intact but completely destroying others. It sent a huge fireball traveling through the south part of the elevator some 40 to 60 feet but northward about 1,200 feet. Structural damage was sustained over the entire length of the building, with large sections in complete or partial collapse. Traditional means of accessing the head house were not available.
Fire department response was delayed momentarily because workers at the facility did not immediately call 911. But it shortly became apparent to the members of the Sedgwick County Fire Department (SCFD), who were scheduled to tour the facility that very morning, that the noise they heard was not thunder from the weather passing through the area.
SCFD responded, confirmed an explosion in the grain elevator, and immediately upgraded the response to a second alarm. Fire department policy forbade immediate entry into the structure for at least five minutes or until such time as the dust could settle so that personnel were not exposed to an imminent secondary dust explosion potential. Chief Gary Curmode established incident command, activated the countywide disaster plan, and took a position in the Emergency Operations Center as appropriate for what obviously was a large-scale incident that would require the resources of numerous county agencies.
SCFD established a “hot zone” perimeter 600 feet from the structure. The Wichita (KS) Fire Department (WFD) responded on mutual aid with one squad and its rescue team. In the first stages of the incident, SCFD had close to 100 firefighters on the scene, including its technical rescue team.
Initial size-up indicated that a blast or series of blasts had blown out both ends of the building. The central head house had large holes blown in it; the first few floors were reduced to rubble. Huge chunks of concrete the size of apparatus were left suspended by a few strands of rebar. The tops of the bins had been dislodged. Damage to the bins was worse on the north end, but the blown-out south end and the pile of rubble indicated obvious structural damage.
There were small spot fires in and around the building–fortunately, these were small and not of major consequence. Water supply was remote, and a shuttle operation was initiated. One and three-quarter inch handlines were stretched to extinguish or contain these small fires. The bulk of the fires were confined to the bins, smoldering fires that at first took possession of four or five bins but eventually extended to 24. It would take a month to completely extinguish these fires.
Within 10 minutes a company representative supplied the on-scene commander with an employee list. Thirty-seven people were unaccounted for initially. With firefighters conducting victim surface removals, this accountability number soon would drop to 12.
Command assessed the rescue priorities and needs. Four employees were thought to be in the head house at the top of the elevator. Others were thought to be in the tunnels. But the condition of the structure and other hazards had to be addressed. Early on, SCFD called in structural engineers and other specialists to assess conditions. One such expert was the architect who designed the building. In addition to these structural checks, the interior atmosphere was constantly monitored for dust, oxygen, and toxic levels. SCFD designated several safety officers to address personnel safety. SCBA were mandated for all interior operating personnel until such time as the atmosphere was deemed safe.
In addition, Command stopped railroad traffic on the nearby line and closed the air space over the structure to minimize the potentially catastrophic effects of vibrations. However, EMS helicopters were called in to transport victims to area hospitals.
Staging areas for emergency apparatus and private contractors were designated and controlled. Early in the incident SCFD requested a Federal Emergency Management Agency (FEMA) disaster response team. An advance party of the Omaha (NE) Disaster Response Team arrived at approximately 2130 hours and the full response unit at approximately midnight of the first night of operations.
By 0945 hours, on-duty Wichita Fire Department rescue team members were called to respond. My team arrived on-scene at 1010 hours.
Members of the rescue companies operated in teams of four. The SCFD technical rescue team was assigned the operational rescue function and my team–Captain Randy Slaughter, Lieutenant Kenny Ast, Firefighter Meredith Dowty, and I–were assigned to reconnaissance. We responded to the south end first. Command directed us to the area in which workers were believed to have been at the time of the explosion. We walked past dump trucks and high loaders that were in operation when the blast occurred. All were extensively damaged: Fiberglass components were shredded, all windows were shattered, and desk-sized pieces of concrete were lying on the equipment. It was evident that the operators had run for their lives; the equipment was still running.
SEARCH FOR LIVE VICTIMS
While working our way to the head house, we came on the first fatality. He had been blown clear of the structure. All his clothing was blown off. It was later determined that he had been thrown some 40 feet by the blast, into a concrete structural member with such force that his body left an imprint where it hit and bounced off. I covered him with a tarp.
Intelligence coming into the command post indicated that possibly four workers had been in the tunnels under the elevator at the time of the explosion. Most of the tunnels were now filled with mountains of rubble and debris. From what we could see, the grain in the bins had been dumped into the tunnels. The grates that had covered the access area to the tunnels were now wrapped around a beam 200 feet above us on the bottom of the catwalk at the top of the elevator–another indication of the force of the blast. We called into the tunnel several times. There was no response. Conditions were such that probable survivability was small.
We were then informed that a live victim had been seen near the head house on top of the structure. The instability of the structure made access to the top an early concern. Command addressed this problem by employing a helicopter and a crane. An Army helicopter requested from nearby Fort Riley provided the command staff with an aerial view of the overall scene. It was also used to remove a victim from the south end of the elevator.
The heavy lifting crane was positioned on the west side of the structure, just north of the head house. Using the bucket as our elevator, the crane brought our team to the top of one of the few still intact north-end galleries. Since the structural stability of the roof there was not certain, we deployed a safety line and our personal harnesses to tie off to each other. There was no secure object on the roof that could serve as an anchor point for our roof operation. One team member was positioned at the front, the second 10 feet behind, and the third and fourth some 40 feet behind the second. The first two rescuers would advance; the other two would meet them.
Moving in this manner, we found our first live victim on the roof of the bins on the east side, some 135 feet above the ground. When he saw us, he started toward us. We instructed him not to move. When we arrived at his location, it was obvious that he was seriously injured. His hands and face were badly burned. We assisted him in reaching the crane basket and helped him to enter it. We now returned to our recon mission.
We could not proceed northward because the tops of the bins had been blown off, and there was no way to pass over the open bins. So we moved south to the head house. We eased through the rubble to the west catwalks. Ast crossed the catwalk and peered into the open doorway. The floor was gone–in fact, two or three floors had collapsed onto another floor. He called out several times. There was no response. We backed out.
We then moved to the east catwalk. Ast again made his way across the walkway to the door. I followed, staying a few feet behind. The floor, again, was gone. We called out for conscious victims. There was no response from the rubble below. We backed out.
Even though the floor on this level was gone, I noticed that the ceiling was intact. I knew we had to get above–it was the only intact area left. We moved outside to the top of the intact bins, from which we had removed the victim earlier. From there, we found a ladder leading to the roof of the head house. Ast, Dowty, and I made our way to the roof. Just above us, the helicopter from Fort Riley was engaged in aerial recon. The vibration from the helicopter shook the entire structure. I asked Slaughter, who was in radio contact with Command, to have the helicopter grounded while we were on the structure.
Two More Victims
From the head house roof, we moved down a ladder to the roof of the catwalk. Ast tied off our rope to the ladder as Dowty and I slowly crossed the catwalk roof. The only entrance to this level was a window about six feet above the catwalk roof. Dowty pulled himself up and called out. This time we heard a reply: “We`re in here!” There was at least one, and possibly more, victims. I relayed this information to Ast, who relayed it to Slaughter, who in turn informed Command. The second rescue team stood by below in rescue readiness.
We spent the next few minutes getting information from the victims. We learned that they were not in the room with the window but were one level up; there were two of them, and both were breathing. But only one was talking.
We told them to stay where they were and that we were on our way. I gave Dowty a leg up, and he entered the building. I requested Ast to come up and assist me similarly. While waiting for Ast to make it to my location, I looked closely at the tower I was getting ready to enter. Large holes were blown out of the east and west walls. The north wall–the one with the window I was preparing to enter, was bulging. Cracks spidered through the entire exterior. Above, two huge chunks of concrete, suspended by rebar, shifted. I remember thinking that if they broke loose, they would wipe out the catwalk–and Ast and me.
Ast gave me a leg up. As I came through the window, Dowty moved to the ladder that ran up the center of the head house. I instructed Ast to stay at the window and relay information to Slaughter, who had the radio. I moved to the ladder and followed Dowty up. It was a tight squeeze through the floor opening to the next level.
The victims were located in a small control room. A door on the east wall led to this room. It was a steel door with two PlexiglasTM panels set in. The top panel was blown out. The victims were close to this east wall, one seated and the other lying down. The exterior west wall of the room was completely gone, and the ceiling sagged in a V-pattern. The explosion had radically deformed the outward-opening door, pushing more than half of it inward past the steel jamb. We could not open it. I asked Ast to request a pry bar and a K-12 saw.
Meanwhile, Dowty used a hammer and screwdriver provided by the victims to try to remove the door pins, hoping to take the door out of the bind. This was unsuccessful. Dowdy then removed the lower PlexiglasTM panel and the metal brace between the two panels, providing an opening to access the victims.
Both victims were seriously injured. Skin was hanging from their burned hands. Their faces were black from burns and smoke. Even though they could not use their hands, both victims thought that they could make it through the window with our help. We placed chairs inside and outside the window and carefully assisted them through.
We would have to use the central ladder in the head house to bring down the victims to the level at which we could access the crane. Neither victim could climb the ladder. Several rescue team members had assembled on the lower floor. I requested that only a bare minimum of rescuers remain below. The structure was very unstable.
Because of the condition of the victims, time was critical. I decided to lower them the quickest way possible. I tied a loop in a 30-foot rope using a figure eight knot with safety. We had the first victim step through the loop, which was brought up under his armpits with the knot at the back of his head. We wrapped the rope twice around the rung of the ladder above the victim`s head and held tension as the victim stepped onto the ladder. As he stepped down, we lowered him. We lowered the second victim in the same manner.
The Final Rescue
Just before we lowered him, the second victim told us he thought he had heard someone on the level above. We free-climbed to the next level, about 50 feet, and encountered another victim. His hands and face also had been burned. His level of consciousness seemed to be somewhat diminished, and his hands were very swollen and dripping profusely.
He could not be removed with a loop as the others. We placed him in a half harness, but he was tipping. He needed a full body harness. Lieutenant Billy Cordts removed his harness and placed it on the victim. Using a carabiner, we hooked to a D-ring behind the victim`s head. By this time, Battalion Chief Dan McClure, rescue commander at this incident, had free-climbed to our level. He stayed below the victim and assisted with the lowering. We made it safely. All four victims–three in critical and one in serious condition–were brought to the ground via crane, where medical personnel waited, and were transported to area hospitals via air ambulance.
I made a quick check of each room on the way out. As I crawled out the window through which we had entered some time ago, I pulled out the stepladder so that no one else would enter this dangerous area.
Once on the ground, I informed Command that the areas we had just vacated were all clear of victims and that no one else should enter there because of the dangerous condition of the structure.
Several workers still were missing, presumed to be in the tunnels. Operations in the tunnels required careful preparation, as conditions there were very dangerous. The blast completely mangled the conveyor system in the tunnels. Huge pieces of steel had been twisted and were sent flying. The grain in the bins was dumped into the tunnels in cone fashion, creating a void space in the tunnel between each bin. Firefighters using harnesses and safety ropes entered the tunnels and swam through the grain searching these void spaces. Engulfment was a major concern; disrupting the cones of grain could allow thousands of bushels above them in the bins to flow on top of them. My assignment was to continuously monitor the integrity of the search ropes as firefighters conducted a systematic search in the void spaces within the hundreds of feet of tunnel. The search continued until the morning of Day 2.
At midnight of Day 1, the Omaha (NE) FEMA Disaster Response Team arrived. At daylight, all rescue efforts were temporarily suspended to allow the FEMA team a chance to assess the situation. My team was relieved at 0700 hours on Day 2, after 18 hours of duty.
The decision was made to use shores to seal the bins and to vacuum the remaining grain out of the tunnels using vacuum railcars. Operations resumed.
Three teams were set up in each tunnel area for the shoring operations: an inside team to take measurements and construct the shore, a backup team positioned at the tunnel opening that also transported the wood to the interior team, and the outside cutting team. Teams were rotated and relieved at regular intervals.
It took more than two hours to construct each shore, and the entire shoring operation lasted six days. The shoring was a two-step process: temporary shoring with screw jacks and oversized 118-inch plywood to secure the 30-inch-square bin covers to stop the flow of grain, followed by permanent shoring underneath it with 4 2 4s.
Firefighters from SCFD, WFD, and FEMA rescue teams worked together. This system worked well, but operations went slowly. It took several days to empty the tunnels. A few problems occurred, including the breaking loose of the shoring holding one of the bins. Several rescuers were cut off by this hill of grain but were rescued quickly. No firefighters were injured at this incident.
Several missing victims were found, deceased, under debris in the tunnels on Days 2 and 3. I assisted in the search for the last missing victim on my return to duty on Day 4. The rescue mode ultimately was suspended before he was found. FEMA was deactivated after a week. The final victim was recovered several days later during the demolition of the elevator.
In all, seven workers died, among them one of the workers we rescued from the head house.
Although we had practiced managing a large and all-encompassing incident under a unified command system, this is the first time we experienced how efficient the system is.
Using a helicopter close to unstable structures such as this one can be dangerous. Its use near the structure was quickly discontinued. Helicopters used for victim transport landed several hundred feet away from the structure.
Anticipate that large-scale incidents will attract nondetailed firefighters looking to help. Before the incident had progressed very far that first day, several off-duty firefighters responded. A system was needed to stage these firefighters and assign them as necessary. A well-maintained staging area helps address this safety issue.
Members from technical rescue units from both fire departments were divided into manageable four-person teams. Each team had a leader and a radio. This enabled us to deploy in an organized manner.
Incidents such as these illustrate the need to bring creativity to the rescue firefighting process. Each operation presents its own unique circumstances and conditions that stand between you and your objective. Be prepared to make changes in your rescue plan. Find creative solutions within appropriate margins of safety as determined by your risk-benefit analysis.
In this incident, the condition of the victims and structural conditions demanded as rapid a removal as possible, therefore a “quick and dirty” method was used over a more technical method of victim removal. Consider the victim condition as part of your ongoing risk-benefit assessment. Don`t be married to the technicality of the rescue if the situation warrants otherwise.
Training in stressful situations provides a built-in safety factor and a certain confidence/comfort level for difficult operations such as this. You play the way you train.
Teams designated for recon should be prepared to switch rapidly to rescue mode if necessary. This includes carrying as much equipment as possible to effect a rescue in a timely manner.
As with any target hazard in your jurisdiction, preplanning–an intimate knowledge of the structure, the nature of the business, its personnel, and associated hazards–is critical to operational success.
Having an emergency disaster plan and utilizing an emergency operations center are important to effective management of large-scale incidents. Practice the plan, and know all the players who will be operating within the unified command system.
Large-scale incidents demand resources. Anticipate resource needs, and call for them early on. Include in your preplanning/notifications list experts who can contribute to your operation. For this incident, SCFD benefited from the knowledge of grain experts, structural engineers, and FEMA USAR personnel.
Tight command and control of all operating forces and support personnel are essential for incidents involving or potentially involving dust explosions. Secondary explosions are of paramount concern, and careful deployment of resources must minimize their exposure to the greatest possible extent. n
There were 16 dust explosions in the United States in 1997. Of these, nine involved grain elevators. There were 64 grain dust explosions in grain elevators between 1988 and 1997. Statistically, corn is the commodity most often involved in explosions, but all grain dusts have the potential to ignite rapidly when suspended in air at the right concentrations. Statistical source: Robert W. Schoeff, Kansas State University.
GRAIN DUST EXPLOSIONS
The reason that dusts burn with explosive quickness lies in their small particle size and the resulting very large surface-to-volume ratio. Therefore, there is an extremely large surface area exposed to air (specifically, to oxygen) when the dust particles are suspended in the air. The materials are thus easily ignitable in suspended dust form, whereas they ignite with difficulty, if at all, in the form of larger pieces. Indeed, the dusts of some products usually considered nonflammable are themselves highly dangerous.
Dust suspensions have lower and upper explosive limits (LELs, and UELs) just as flammable vapors do. The meaning of the terms is also the same when applied to dusts. LEL is the minimum concentration of suspended dust necessary to sustain flame propagation. Above the UEL, the dust concentration in air is too great for sustained combustion to occur.
It is common to speak about dust “explosions,” but this is not always accurate. Some dusts deflagrate rather than detonate. A deflagration occurs when the flame speed through the gaseous combustion products is less than the speed of sound (1,088 feet per second). A detonation, with its accompanying shock wave, occurs when the flame speed equals the velocity of sound. Brisance is a property of the shock wave from a detonation and is a measure (psi) of its shattering effect.
A dust explosion can occur only when the dust is suspended in air; a dust fire occurs in deposits of dust on flat surfaces. Don`t leap to the false conclusion that a dust fire is of less concern than an explosion, though. The transition from a fire to an explosion can take place virtually instantaneously. The only requirement is that the dust deposit be disturbed suddenly by a heavy hose stream, for example.
Fire can also follow rather than precede a dust explosion when the force of the explosion scatters burning particles, igniting dust accumulations throughout the room or building.
Another possible sequence is that of a primary or initial explosion, followed by secondary explosions. The first explosion, usually relatively minor, will cause the accumulated dust from various surfaces in the area to become suspended. Burning particles from the primary explosion then ignite the newly suspended dust and a second, and usually more severe, explosion occurs. The series of explosions can continue through a third and more, each progressively more severe.
One of the greatest dangers from a dust explosion is the expansion effect, which leads to a great pressure increase in a restricted area. Grain dust explosions can generate pressures of 150 psi and higher. When you consider that many buildings are designed to withstand no more than two psi of extra lateral pressure, the total structural collapse that often accompanies a grain elevator explosion is readily understood.
The rate of pressure rise is also important as far as structural damage and personal injury are concerned. A brisance exceeding 15,000 psi per second is frequently reported, a level that is likely to be fatal to any human in the vicinity.
The temperature of the flame front that carries through a room or building as a result of a grain dust explosion is important for at least two reasons. First, many of the serious injuries and deaths are due to burns. Second, fire is quickly spread throughout the facility by a high-temperature shock wave, igniting all combustibles in its path.
Don`t overlook the possible presence of toxic gases and combustion products after a dust explosion. This is a particularly serious problem for employees who are likely to be injured and rendered unconscious for at least a few moments after an explosion. Being unable to evacuate themselves from the area, these people can very quickly inhale lethal combustion gases. No firefighter should ever enter any building where an explosion has occurred without SCBA.
The challenge of dust fires is actually twofold: to extinguish the fire with minimal damage and to prevent an explosion.
Fires in dust deposits can be either smoldering or flaming fires. The smoldering fire will often be found deep within a pile of dust and may have been burning for a long time before it was discovered.
The greatest danger to firefighters attempting to extinguish a dust fire lies in the rapidity with which the fire can be transformed into a dust explosion. The dust must absolutely not be disturbed in any way that would cause it to become suspended in the air. Remember, the ignition source is already present. The development of the large surface-to-volume ratio is all that is needed to generate an explosion.
The safest course of action could be no action at all, just let the fire burn itself out. The circumstances in which this is practical are obviously limited; however, and a constant watch will have to be maintained.
More often, the fire will have to be attacked and extinguished. The type of dust involved will dictate the extinguishing agent that is used. Water can be used on all grain dusts and, in fact, on just about any dust except those of combustible metals (class D fires).
Remembering the critical importance of not suspending any dust in the air, use fog lines only. And, even with a fog stream, operate it at the lowest practical pressure. Never use a straight stream or high-pressure fog on any dust fire. To do so will only invite an airborne dust cloud and explosion.
Another caution should be mentioned. Do not walk on a pile of burning dust. Your body weight could cause collapse of surface layers covering hollow pockets that have burned underneath.
–“An Overview: Grain Dust Explosions,” John E. Bowen, Fire Engineering, May 1983.
BILLY JACK WENZEL is a captain and certified EMT in the Wichita (KS) Fire Department, where he has served since 1981. He formerly was a member of the department`s hazardous materials team. In addition to N.A.U.I. scuba diver certification and various others, he is certified in the following areas: advanced high angle rescue, confined space rescue, water rescue, trench rescue, vehicle extrication, ice rescue, ventilation techniques, advanced crash rescue, basic water rescue, and fast water rescue. He has an associate`s degree in fire science.