PRESSURE EFFECTS on AND DEFORMATION of WASTE CONTAINERS
MICHAEL D. LARRAÑAGA
CFPS
DAVID L. VOLZ
FRED N. BOLTON
PE, CIH
More than 123 incidents involving incompatible chemical mixing and pressurization of drums in which the drums became pressurized to unknown levels have occurred within the Department of Energy since 1992.1 Failure of a pressurized metal drum can cause a rapid release of pressure, which could cause the drum top or bottom to eject. The resulting flying debris and release of the drum`s contents could injure personnel and con-taminate the environment. Personnel can also be exposed to radioactive or hazardous contents and pyrophoric, flammable, or combustible materials, which can ignite and burn.2
Some hazardous-materials teams have little or no training in how to respond to or approach bulging drums and may not be aware of the many unknowns when working with pressurized drums. Personnel, for example, may not know the internal pressures of the drum or its degree of stability, how to remove the bung or puncture the lid to relieve the pressure, or whether the drums should be approached at all. Yet, pressurized drum incidents occur frequently across the United States–in the public, private, and government sectors, as can be seen from the following examples:
In April 1996, the Lodi (CA) Hazardous Materials Team responded to two 55-gallon metal drums dumped on the property of a commercial restaurant. It was determined that one of the barrels was out of round, bulging, and pulsating due to pressurization. The haz-mat team concluded that approaching the drum was extremely hazardous and used a police bow marksman to shoot an arrow at the drum to relieve the pressure. The mission was completed with one arrow, and the drum contents proved to be pH14 and waste from a clandestine drug lab in another part of the city.3
The Albuquerque (NM) Fire Department, in the summer of 1997, responded to a bulging drum and used a police marksman to shoot a bullet at the drum to relieve the pressure. The mission was accomplished with one bullet. It was later determined that the drum contained acetone, a flammable organic material.
In July 1997, at the Brookhaven National Laboratory, a metal lid from a pressurized eight-gallon container blew off, striking an operator in the face when she attempted to loosen the retaining lid.
Currently, there is no quick, inexpensive, and reliable method for determining pressures inside drums.
LOS ALAMOS RESEARCH OBJECTIVES
In an attempt to develop training criteria and a device for determining internal pressures in 55-gallon drums, the Los Alamos National Laboratory conducted research to explore the effect of pressure on new closed- and open-head 55-gallon metal and plastic drums, 30-gallon metal and plastic drums, 20-gallon plastic drums, and 85-gallon metal overpacks.
The objectives of this research were the following:
to determine the pressure at which 20-, 30-, and 55-gallon drums (metal and plastic) and 85-gallon metal overpacks fail;
to quantify the amount of deformation 55-gallon metal drums experience at various pressures under separate treatments;
to determine if the data for the 55-gallon metal drums support developing an instrument for determining internal pressures;
to conduct a statistical analysis on the mean failure pressures of the collected data for 55-gallon drums.
The information gained was to provide critical information to emergency responders, including tools for assessing a bulging drum situation. The most noteworthy of these tools is the “safe working envelope,” which will help responders determine how to deal safely with and approach pressurized drums.
MATERIALS AND METHODS
Drums of different types were pressurized from zero pounds per square inch gauge (psig) to failure at 5.0 psig intervals. Linear deformation along the center line of metal drums was measured and recorded with corresponding pressures. A test apparatus was built to complete measurements and allow for pressurization. A containment cage was built around the apparatus to contain flying debris ejected during drum failure.
Pressure was increased for each test interval and then allowed to stabilize (less than 30 sec) until both gauges were equal. When measuring deformation, measurements were taken with two devices placed in the center of the top and bottom ends of each drum. Both devices consisted of a rod that traveled up with increasing deformation and indicated a graduated measurement in a 1:1 ratio. Both the top and bottom devices were zeroed before each test.
Observations were made through a 30 x 60 magnification spotting scope approximately 75 feet from the test apparatus. Measurements were clearly visible, and zeroing of the gauges was completed by adjusting the gauges and linear measuring devices and confirming zero through the spotting scope. The reason for collecting data remotely was to guard against the potential dangers of flying debris and rapid pressure release during the pressurized drums` failure.
INDIVIDUAL DRUM-TYPE TESTS
All drums tested were new and unused to eliminate the bias that could be associated with drawing conclusions from old or used drums.
55-Gallon Metal Drum Tests. Two separate metal 55-gallon drum types were used in this experiment: the UN/1A1 closed-head and the UN/1A2 open-head. The UN/1A1 closed-head drums were pressurized and observed under three separate treatments: (1) one-half full of water; (2) three-quarters full of water; and (3) empty. The UN/1A2 open-head drums were pressurized and observed under three separate treatments: (1) one-half full of water; (2) three-quarters full of water; and (3) cement spun (partially filled with cement and spun in a machine similar to a centrifuge to simulate radioactive waste packaging techniques at Los Alamos National Laboratory). The rings of the empty and water-fill, open-head drums were tightened using an impact wrench and sledgehammer to within one centimeter of the point where the ring ends meet. The rings of the cement-fill, open-head drums were tightened to 40 pounds of torque using an impact wrench.
Each drum type and respective treatment was pressurized from psig to a failure pressure at 5.0 psig intervals. Linear deformation along the center line of the metal drums was measured and recorded with corresponding pressures. The purpose of the treatments was to determine if differences in deformation exist between drums of each treatment and between drum types and if mean failure pressures were significantly different.
55-Gallon Plastic Drum Tests. Five 55-gallon seamless, high-density polyethylene (HDPE) drums were tested to determine their failure pressures. Deformation of these drums was not measured because both radial and axial deformation were observed. Four of the 55-gallon HDPE drums were tested empty, and one was tested one-half full of water. These drums were steadily pressurized until failure. Two open-head HDPE drums were tested to determine their fail pressures. One plastic-lined metal drum was pressurized; deformation measurements were recorded up to failure.
85-Gallon Metal Overpack Tests. Six overpack drums were tested empty to de-termine a mean failure pressure for the drums. Top deformations were measured.
30-Gallon Metal Drum Tests. Four drums–two closed-head and two open-head–were tested to determine their failure pressures. Deformation was not measured because the measuring devices were not designed to measure deformation of 30-gallon drums. The drums were slowly pressurized until failure occurred or the pressurization was discontinued because of dangerously high pressures.
20- and 30-Gallon Plastic Drum Tests. Two 20-gallon seam construction and four 30-gallon seamless HDPE drums were pressurized to determine their failure pressures and characteristics.
PROGRESS AND RESULTS
The tests revealed the following differences in the failure characteristics among drum types.
55-Gallon, Metal, Open-Head Drums.
–The drums appear to vent immediately adjacent to the nut-and-bolt fastener on the ring.
–All drums tested vented at pressures at or below 32 psig.
–Pinging was noticeable between 15 and 20 psig.
–The drums appeared to bulge at only the top and bottom ends.
–Body seams (top to bottom) experienced no visible distortion or apparent weakening.
55-Gallon Metal Closed-Head Drums
–95 percent of the drums tested failed explosively.
–Of the catastrophic failures, 68 percent failed at the bottom end, making the entire drum a projectile.
— All drums tested failed at the top or bottom ends.
–When filled with liquid (12 or 34 full), bottom failures appear to be increasingly violent with increasing water levels up to 34 full.
–Approximately 5 psig before failure, a significant amount of distortion of the drum chime (rim) is apparent (see Figure 3 on page 60).
–The 55-gallon metal open-head drums appear to bulge at only the top and bottom ends.
–Body seams (top to bottom) experienced no visible distortion or apparent weakening.
–Pinging was noticeable between 15 and 20 psig and increased dramatically immediately before drum failure.
–Statistical analyses of the test results indicate a probability that 99 percent of the failures will occur above 48.7 psig.
These observations show that bulging drums, especially the closed-head type, are extremely dangerous. Noticeable differences exist between closed- and open-head drums under pressure. However, both types of drums are inherently dangerous when pressurized and present many hazards. The testing provides a reasonable certainty that new closed-head 55-gallon drums will fail above 48.7 psig and helps in determining a “safe working envelope” for working with these drums.
The averages of the top deformation of the closed- and open-head drum treatments were plotted against pressure (see Figure 4 on page 60). The open- and closed-head top deformation averages are sufficient to use the data in developing a device to correlate pressure vs. deformation for estimating the internal pressures of visibly bulging 55-gallon metal drums.
55-Gallon Plastic Drums. The five 55-gallon HDPE-seamless drums tested failed explosively at pressures of 48, 48, 50, 30, and 58 psig. Four of five failures occurred through the sides of the drums at no particular or identifying location. One failed at 30 psig out of the top end of the drum. These are significant observations because they show the potential for seamless HDPE drums to fail out of the sides. Deformation was observed at the tops, bottoms, and sides of the drum.
The two 55-gallon, open-head HDPE drums failed explosively at 23 and 24 psig, ejecting the entire top off the drum. One 55-gallon plastic-lined metal drum self-vented at 50 psig, with a top bulge characteristic of the curve for the closed-head top deformation. A device for estimating internal pressures in 55-gallon drums will not be useful on plastic or plastic-lined drums because deformation was not measured or construction differences exist between these drums.
85-Gallon Metal Overpacks. The six overpacks tested failed at or below 16 psig and appeared to self-vent immediately adjacent to the placement of the nut-and-bolt fastener on the ring. The overpacks appeared to bulge only at the top and bottom ends.
30-Gallon Metal Drums. The four drums (two open-head, two closed-head) tested showed that significant hazards exist when a 30-gallon metal container is pressurized. The high pressures associated with these containers present many hazards.
–30-Gallon Metal Closed-Head Drums:
extremely high pressures possible (>120 psig),
extremely high pressures maintained without venting, and
catastrophic and extremely violent failure is anticipated.
Other than bulge, no apparent failure indicators, such as pinging, were noted. The drums appeared to bulge at only the top and bottom ends.
–30-Gallon Metal Open-Head Drums:
Of the two drums tested, one failed explosively and one self-vented.
Both drums tested maintained pressures greater than 50 psig.
Other than bulge, no apparent indicators, such as pinging, were noted; the drums appeared to bulge at only the top and bottom ends.
20- and 30-Gallon HDPE Plastic Drums. The tests revealed the following:
The seam and seamless construction drums failed at pressures above 45 psig.
Both the seam and seamless construction drums can maintain high pressures for extended periods of time.
Two of four seamless construction drums failed explosively from the side at no particular location on the drum, making the entire drum a projectile.
The seam construction drums failed explosively along the bottom seam of the drum, making the drum a projectile.
The drums appeared to bulge from the top, bottom, and sides.
CONCLUSIONS
Deformation of a drum indicates that the drum has been subjected to internal pressure, not that it is under pressure at the time of inspection. The container`s design makes it capable of violent rupture. Always approach deformed drums with caution. It was assumed that the experimental method of pressurization at increasing 5-psig intervals was representative of “real” pressurized drums in the field, although the speed and method of pressurization may be different. Also, it was assumed that the deformation observations correspond to “real-world” pressurized drum deformations in the field.
The results of this study indicate that significant differences exist among the failure characteristics of drum types and materials. The 55-gallon drum data were sufficient to support the development of a device for estimating internal pressures in 55-gallon metal drums because of the similarity between pressure vs. deformation curves. The safe discovery and mitigation of a pressurized drum can make the difference between a minor incident with limited and controlled consequences and a major one with the potential for loss of life and significant property damage.4 With the information contained in this report, Department of Energy personnel, private and municipal fire departments, haz-mat teams, hazardous-device teams, and other emergency responders can make educated decisions during bulging drum incidents. Many observations were noted and should be used as indicators of danger when approaching bulging drums.
INDICATORS AND OBSERVATIONS
The following indicators and observations are considered to be of extreme importance for emergency responders and waste workers:
The chime on 55-gallon, metal, closed-head drums distorts at approximately 5 psig before failure.
Intermittent pinging of 55-gallon, open- and closed-head metal drums occurs at 15 to 25 psig.
Rapid and intense pinging of 55-gallon, metal, closed-head drums occurs immediately before drum failure.
A strong potential exists for closed-head 55-gallon and closed- and open-head 30-gallon metal drums to fail explosively, making the entire drum a projectile.
All the 85-gallon and 55-gallon open-head drums tested self-vented.
30-gallon, metal, closed-head drums can hold and maintain in excess of 120 psig.
One of two open-head 30-gallon metal drums tested failed explosively.
Four of five seamless 55-gallon closed-head HDPE drums failed explosively out of the side of the drum.
Two of four 30-gallon HDPE drums failed explosively out of the side of the drum, and two self-vented.
The seam construction 20-gallon drums tested failed explosively on the bottom seam.
Sealed containers of all types have the potential to become pressurized.
RECOMMENDATIONS
Based on the conclusions of this study, we recommend the development of training criteria using these data. The training criteria should include the following information:
indicators of drum pressurization;
inherent hazards associated with drum incidents and chemical properties such as flammability, corrosiveness, and reactivity;
failure characteristics and differences between types of drums; and
pressurized drum-mitigation techniques such as remote venting, direct cooling (ice bath), or shooting with a projectile (water cannon or disrupter).
 |
Pressure Effects on and Deformation of Waste Containers
 |
 |
Figure 2. 55-galloon drum (left) and 85-gallon drum (right). Open-Head drums leak immediately adjacent to the nut-and-bolt fastener on the ring. Note the crease in the metal at that location. (Photos by Michael D. Larrañaga, CFPS.)
 |
 |
Figure 3. The chime (rim) of the 55-gallon, closed-head, metal drums becomes severely distorted immediately before catastrophic failure (right).
 |
References
1. Operating Experience Weekly Summary 97-32, Office of Nuclear Safety, Washington, D.C., U.S. Department of Energy, Aug. 1-7, 1997.
2. “Fire, Explosion, and High-Pressure Hazards Associated with Waste Drums and Containers,” Safety Notice, Issue No. 93-1, Office of Nuclear Safety, Washington, D.C., U.S. Department of Energy, Feb. 1993.
3. Howard, Hank A., “Clandestine Drug Labs: An Epidemic in the United States,” Fire Engineering, Sept. 1997.
4. Plutonium Storage Safety at Major Department of Energy Facilities, Defense Nuclear Facilities Safety Board Technical Report, U.S. Department of Energy, April 14, 1994.
For additional information about the waste container report, e-mail Michael Larrañaga at
MICHAEL D. LARRAÑAGA, CFPS, is a health and safety engineer in the Industrial Hygiene and Safety Group at Los Alamos National Laboratory and a team member of the Department of Energy`s worldwide deployable accident response group. He has a master`s degree in industrial hygiene and safety from the University of Houston-Clear Lake and a bachelor`s degree in fire protection and safety engineering technology from Oklahoma State University and is currently pursuing postgraduate work in chemical engineering at the University of New Mexico.
DAVID L. VOLZ is haz mat specialist-coordinator, Hazardous Materials Training Center (TA-49) Hazardous Materials Response Group, Los Alamos National Laboratory. He organized the Hazardous Materials Training Center (TA-49) with equipment and materials from salvage yards, accidents, and donations. He conducts haz mat training throughout the Navajo Nation for the New Mexico Water and Waste Water Association, Department of Energy, and local fire departments.
FRED N. BOLTON, PE, CIH, is ESH team leader in the Nuclear Materials Technology Division, Los Alamos National Laboratory and a team member of the Department of Energy`s Radiological Assistance Program and worldwide deployable Accident Response Group. He has a bachelor`s degree in safety engineering and a master`s degree in industrial hygiene from Texas A&M University.
