HAZARDS ASSOCIATED WITH RETRIEVAL AND STORAGE OF
LEGACY WASTE AT THE TRANSURANIC WASTE
INSPECTABLE STORAGE PROJECT
Phillip W. Grogin, CHMM, Michael A. Pannell, CIH, Rhonda Langford, Gilbert Montoya.
Environmental Management-Solid Waste Operations
Los Alamos National Laboratory
Los Alamos, New Mexico, 87545
Approximately 16,900 containers of solid transuranic and hazardous waste have been stored beneath earthen cover for nearly twenty years at Technical Area 54 of the Los Alamos National Laboratory. The mission of the Transuranic Waste Inspectable Storage Project (TWISP) is to retrieve, vent, and place these containers into an inspectable storage configuration in compliance with the Resource Conservation and Recovery Act, prior to final disposition at the Waste Isolation Pilot Plant. Significant hazards currently identified with TWISP activities include: 1) the pressurization of drums; 2) the emission of volatile organic compounds (VOCs) from drums; and 3) the generation of elevated hydrogen levels by certain waste streams. Potential causes of increased drum pressure include extreme temperature variations, radiolysis, and other chemical reactions. Pressurized drums pose physical, chemical, and radiological hazards and increase the chance of contamination to both personnel and the environment. Visual inspection for distended drum lids and audible identification using acoustic pitch and resonance are the primary methods of detecting elevated drum pressures. During the processing of potentially pressurized drums, risks are reduced by utilizing control measures such as tagging suspect pressurized drums and prioritizing these drums for venting. The presence of VOCs may be inherent to the source waste stream or generated by biological or chemical reactions. VOCs have been identified in container headspace samples. Analysis of drum headspace samples has confirmed notable concentrations of methylene chloride, acetone, methyl ethyl ketone, benzene, toluene, and ammonia. These compounds may pose inhalation hazards to personnel inspecting and handling vented drums. While VOC concentrations within some drums have exceeded permissible exposure limits by factors of seven or more, breathing zone concentrations have yet to exceed action levels. Elevated hydrogen levels may be attributed to alpha hydrolysis of hydrogenous waste matrices. Analysis of gas samples from vented drums has verified hydrogen concentrations exceeding six times the lower explosive limit in certain waste streams. Reducing the risk of explosion involves monitoring for hydrogen, controlling ignition sources, and nitrogen purging during venting operations. These hazards, combined with the inherent risks encountered at a hazardous waste operations site, present significant challenges for environmental, health, and safety (ES&H) professionals and hazardous waste managers. Based on the retrieval of 15 percent of the waste containers, the following preliminary conclusions are presented to better protect personnel and the environment. 1) The likelihood of unvented drums becoming pressurized increases when ambient temperatures increase. 2) Pressurized drums must be vented before they become bulging drums. 3) Vented drums present the potential for personnel exposure to VOC emissions. 4) Large numbers of co-located vented drums may present the potential for increased hydrogen and VOC concentrations within unventilated storage domes. 5) Monitoring and sampling within domes where vented drums are stored is necessary to ensure that the levels of risk to drum handlers and inspection personnel are acceptable. 6) Identifying, tagging, and prioritizing or segregating special case drums is necessary to prevent unacceptable personnel exposures and preclude environmental contamination. 7) Acoustic drum pressure detection may be a viable tool in assessing elevated drum pressures.
The Transuranic Waste Inspectable Storage Project (TWISP) is an effort to safely retrieve approximately 16,900 containers of transuranic (TRU) and mixed waste from earthen-covered storage and place them into an inspectable storage configuration. The TWISP was initiated in response to a New Mexico Environment Department Compliance Order under the New Mexico Hazardous Waste Act to bring container storage into compliance with New Mexico Hazardous Waste Management Regulations. The TWISP site is located at Area G of Technical Area 54 (TA-54-G) of the Los Alamos National Laboratory (LANL) in Los Alamos, New Mexico.
Both TRU and low-level radioactive waste have been stored in shallow land burial sites at TA-54, Area G since 1957. In 1970, the Atomic Energy Commission directed its facilities around the country, including LANL, to begin storing TRU waste in a manner that would allow for eventual retrieval. Retrieved waste would then be packaged for shipment to the Waste Isolation Pilot Plant (WIPP), a deep geologic repository in southeastern New Mexico. The waste was stored in metal drums and fiberglass-reinforced, plastic-coated plywood (FRP) crates. The drums and FRP crates were placed in densely packed arrays, referred to as waste stacks, on aboveground, earthen-covered asphalt storage pads. Between 1979 and 1991, LANL constructed and stacked three aboveground storage pads with a total of 16,641 drums and 187 FRP crates.
The TRU waste stored on the three pads is primarily comprised of radioactively contaminated materials in solid form. About 60 percent of the waste has been identified as TRU mixed waste. Resource Conservation and Recovery Act (RCRA) constituents such as lead, chromium, beryllium, and spent halogenated solvents are present in some of the waste streams, as may be asbestos. The FRP crates primarily contain individual gloveboxes or portions of gloveboxes and radioactive waste transfer lines from decommissioning projects. Although RCRA hazardous substances comprise only a small fraction of the total waste volume, that fraction is distributed throughout a large number of waste containers.
Pad 1, where retrieval operations began, is comprised of 27 waste streams. Forty-four percent of the drums contain vacuum filter cake, a dewatered sludge generated by the vacuum filtration of solids from pretreated aqueous waste slurry. Twenty-three percent is combustible waste, including paper, rags, plastic, rubber, and cellulose-based waste generated in glove box operations. An additional twenty-three percent of Pad 1 is characterized as noncombustible waste including items such as small tools, cans, small equipment items, glass, and metal waste. Approximately ten percent of Pad 1 drums contain a combination of combustible and noncombustible waste. Waste streams are designated according to their source location; first by technical area, then by operation. Table I summarizes Pad 1 waste streams.
Table I. Pad 1 Waste Stream Summary
DESCRIPTION OF OPERATIONS
TWISP work activities include overburden removal, waste container retrieval and transport, and drum washing and venting prior to placement in an inspectable storage configuration. After several years of planning and constructing retrieval and storage domes, retrieval operations began on Pad 1 in March of 1997. A temporary retrieval dome was constructed over the pad to provide environmental containment during the retrieval process. The retrieval dome is equipped with a High Efficiency Particulate Air (HEPA) filtered ventilation system designed to maintain a negative pressure work area during retrieval operations. Retrieval begins with overburden removal, where soil is removed from one end of the waste stack, called the working face. As the working face of the stack retreats during retrieval operations, the remaining soil is removed as necessary. Following a visual inspection for integrity, drums are removed from the stack one at a time, surveyed for radiological contamination, and then loaded for transport to the Drum Preparation Facility. Damaged or corroded drums are immediately overpacked into larger, vented metal drums. FRP crates are inspected for integrity, surveyed for radiological contamination, and taken directly to storage. Externally contaminated FRP crates are decontaminated or repackaged. Damaged FRP crates are overpacked or the contents are transferred to another container.
At the Drum Preparation Facility, retrieved drums are placed on drum dollies and rolled to a wash bay, where all drum surfaces are cleaned with an environmentally safe solvent. Drum washing is required in order to remove the rust inhibitor with which the drums were coated prior to being stacked and buried. With the rust inhibitor removed, another inspection for drum integrity is conducted. If no problems are identified, the washed drums are then rolled to the Drum Venting System (DVS).
The DVS was designed and constructed at LANL specifically for the TWISP. The DVS is built around an American Society of Mechanical Engineers (ASME) approved stainless steel containment vessel (CV) which is rated to 50 pounds per square inch and is equipped with a HEPA filtration system to provide control of potential particulate and radiological contamination. In addition to the physical barrier it provides, the CV is designed to contain a drum burn, defined as an ignition of flammable components within the drum.
At the DVS, each drum is punctured to release gases that may have accumulated during storage. Then, a vent containing a sintered metal filter is installed using a vent insertion device located within a glove box mated to the top of the CV. The vent allows the release of gases during storage while preventing the release of particulate. During insertion of the vent, a sample of the drum headspace is obtained and the hydrogen content and drum pressure are determined. Drums with a hydrogen gas content above 4% are purged with nitrogen gas before the vent is fully inserted. Once vented, the drums are again surveyed for radiological contamination before being transported to inspectable storage domes. All drums will be analyzed using real time radiography and a representative number of drums will be randomly sampled to verify waste stream characteristics before shipment to WIPP.
TWISP HEALTH AND SAFETY HAZARDS
Although physical hazards such as falling drums, pinch points, and vehicle and heavy equipment operations present the greatest likelihood for personal injury, several uncommon health and safety hazards have been identified at the TWISP. These hazards may apply to other large-scale drum recovery operations and hazardous waste storage facilities, and therefore are not unique to the TWISP. The hazards presented are pressurized drums, VOC emissions from vented drums, and hydrogen generation within drums.
Pressurized drums present a variety of hazards to TWISP personnel, both before and after venting. Prior to venting, increased drum pressure may cause a drum to violently eject its lid. Personnel in the vicinity may be physically injured by the ejected lid and/or drum contents, and may be exposed to chemical and radiological contamination. Environmental contamination may also result. After the drum is vented, gases may continue to be generated and emitted from drum vents, thereby potentially exposing personnel and/or the environment.
Unfortunately, pressurized drums are often difficult to accurately identify. Several cases of drum lids being ejected from pressurized drums have been documented within the DOE complex (1,2,3). In one example, during routine waste sampling operations at LANL Technical Area 3, a lid from a drum that had recently been packaged violently ejected into the air. The drum had been packaged in accordance with LANL guidelines and showed no visible indication of pressurization (4).
Since the TWISP must retrieve approximately 16,900 waste containers, even a small percentage of pressurized drums may present a serious health and safety hazard. As of October, 1997, DVS drum pressure data was available for 323 drums. Twenty-five of these 323 drums had pressures in excess of 3 psig (pounds per square inch gauge): fourteen drums were from waste stream 50-19 (vacuum filter cake), five from 55-19 (combustibles), and 6 from miscellaneous waste streams. Of the 4816 drums on Pad 1, 482 are projected to have internal pressures exceeding 3 psig or greater. Figure 1 presents the projected percentage of pressurized drums by waste stream. This projection would estimate 1700 drums within the entire TWISP drum population to be pressurized to 3 psig or greater.
Fig. 1. TWISP Drums Projected to Contain Elevated Hydrogen and Pressure Levels
Drum pressurization may be caused by chemical, radiological, and biological reactions, and is affected by ambient environmental conditions. Chemical reactions between waste stream constituents may readily generate volatile components. Physical properties of the waste constituents, such as vapor pressure and boiling point, affect the initiation and progression of chemical reactions, especially when external ambient temperatures are increased. Elevated ambient temperature has been shown to increase pressure within unvented drums which have been removed from the cool retrieval pile and placed into hot storage domes. In some cases, even empty waste containers may become pressurized due to increased temperature and barometric pressure. Radiolysis is a proven source of hydrogen generation in waste streams with radioactive constituents, and is also considered a potential source of drum pressurization. Biological reactions as a source of increased drum pressure are processes not well understood; however, bacterial denitrification has been implicated in the generation of ammonia in TRU waste containers (5).
At the TWISP, drums believed to be pressurized are categorized as suspect pressurized or visibly bulging. A suspect pressurized drum is one in which the drum lid center appears higher than normal, but below the plane of the drum ring, or a drum lid in which acoustic testing indicates an elevated pressure. A visibly bulging drum is defined as a drum in which the center of the drum lid extends above the plane of the drum ring, a drum that has a distended body or base, or both. In most cases, the first visible indication of a pressurized drum is a bulging lid. As drum pressure increases, the drum base and body become distended. Distended drums are a major concern, as test drums pressurized in excess of 25 psig have shown minimal distension. If a visibly bulging drum is observed, operations in the area are halted, ES&H personnel are notified, and the drum is left in its original location. TWISP ES&H personnel inspect the drum to determine the category of pressurization. Drums suspected of being pressurized are tagged, and in extreme cases of pressurization, ES&H personnel determine whether the drum presents an imminent hazard. EM&R/HazMat personnel are called in to remotely vent drums considered imminently hazardous. Remote venting involves segregation, remote penetration, and vent insertion. Other drums deemed safe to handle are immediately transported to the DVS for venting
ES&H personnel use acoustic testing or a drum pressure detector (DPD) to determine whether a drum is actively pressurized. Acoustic testing consists of tapping the drum lid and comparing the duration and frequency of the resonant ring to that of a known unpressurized drum. Typically, if the duration is shorter and the frequency is higher, the drum is considered to be suspect pressurized. The DPD is an unproven, experimental approach to drum pressure determination that shows great potential as a tool to assess drum pressurization. The DPD utilizes a magnetic transducer and a spectral analyzer to test for drum pressure. One problem identified with the DPD is the dampening and attenuation of sound resonance when the drum lid has corrosion and/or has a portion of the waste contents touching the drum lid. Research is underway to make the DPD a viable indicator of drum pressure.
Volatile Organic Compounds
The potential exists for worker exposures to chemicals that may be present in waste containers. The primary source of chemical exposures to TWISP personnel is the inhalation of VOCs emitted from vented drums. Acute exposures to VOCs may cause dizziness, nausea, and vomiting. Chronic exposures to certain VOCs may cause central nervous system depression, peripheral neuropathies, and possibly cancer.
Anticipation of personnel exposures to chemical hazards is difficult due to poor waste stream characterization. Historically, TRU waste characterization focused primarily on radiological constituents. Chemical constituents within a waste stream may have been overlooked during waste stream characterization and may not be included in waste profile data. Air samples collected in 1994 from a mixed waste storage dome at TA-54-G provide useful data on the type and concentration of VOCs that may be expected during the TWISP. Compounds detected in these samples include benzene, vinylidene chloride, methylene chloride, and carbon tetrachloride, all of which are classified by the American Conference of Governmental Industrial Hygienists (ACGIH) as confirmed or suspected carcinogens. Drum headspace analysis has proven the presence of hazardous VOCs in concentrations exceeding their respective action limits.
TWISP ES&H personnel perform VOC monitoring and sampling to ensure that worker exposures are kept below occupational exposure limit action levels and as low as reasonably achievable, where possible. An action level of one-half the permissible exposure limit (PEL) or threshold limit value (TLV), which ever is lower, is established for all hazardous compounds. PELs are legally enforceable exposure limits established by the Occupational Safety and Health Administration and TLVs are recommended exposure guidelines established by the ACGIH. Comparison of sampling and/or monitoring data to action levels is used to determine whether procedural changes and/or additional control measures are required.
Direct read instruments, such as photoionization detectors (PIDs), are useful in determining whether sampling is necessary. PID monitoring in drum storage areas has shown concentrations ranging from <1 to 126 parts per million (ppm) at drum vents. In most cases, these measurements drop to background one foot away from the vent. Background is the ambient airborne concentration of VOCs which is expected to be naturally present or is the result of some activity which uses or generates volatile compounds. Usually background is zero, indicating no presence of VOCs. During drum washing, however, PID area monitoring has indicated background VOC concentrations ranging from 5 to 40 ppm. The high background concentrations are due to the use of the drum cleaning solvent which masks the detectability of compounds with low PELs or TLVs. In areas with elevated VOC backgrounds, sampling for compounds with action levels lower than the background concentration is performed more frequently.
Industrial hygiene samples are collected whenever PID measurements indicate the likelihood of personnel exposure, when new operations are initiated, or whenever ES&H personnel feel it is necessary. Samples are also collected in an effort to characterize VOC exposures during specific TWISP activities. As of October, 1997, more than 94 area and personal VOC samples had been collected. Analysis of VOC samples has indicated the presence of methylene chloride and carbon tetrachloride in the breathing zone of workers in storage domes. None of the samples indicated personnel exposures had exceeded action levels. Analysis of headspace gas samples collected from three visibly bulging drums vented remotely by EM&R/HazMat during the summer months revealed concentrations of 9 ppm benzene, 40 ppm methylene chloride, and 15,000 ppm methyl ethyl ketone. The three drums were from the same waste stream (55-38), indicating that certain waste streams have a propensity to emit VOCs under conditions of increased temperature. Continued sampling and monitoring is necessary to ensure worker exposures are maintained at acceptable levels.
Elevated levels of hydrogen gas in drum headspaces have been routinely detected in waste drums retrieved during the TWISP. Initial estimates of hydrogen-producing drums suggested that only one hydrogen-producing drum would be found among all Pad 1 drums (6). Hydrogen concentrations measured during venting operations indicate initial estimates of the number of hydrogen-producing drums were very low.
The generation of hydrogen within TRU waste-containing drums may result from alpha radiation induced hydrolysis of hydrogenous waste matrices or galvanic reactions of dissimilar metals (7,8). Drums with tightly sealed lids that are stored for extended periods of time may build up elevated hydrogen concentrations. High concentrations of hydrogen in unvented drums represent an imminent hazard to personnel due to the potential of explosion. Lower concentrations of hydrogen gas escaping through drum vents may create a potential explosion hazard if a sufficient number of drums are co-located in unventilated storage domes.
Hydrogen monitoring is conducted during drum venting and once the drums are in storage. Unfortunately, difficulties in obtaining a proper seal between the drum lid and the DVS gasket have limited the number of drums for which reliable headspace data have been obtained. As of October, 1997, hydrogen concentrations measured by the DVS were obtained for 323 drums, from a total of approximately 2500 vented drums. Concentrations measured during drum venting have been as high as 23.97%. In one instance when EM&R/HazMat was called to remotely vent a visibly bulging drum, subsequent analysis of the headspace gas revealed a hydrogen concentration of 27.89%.
Hydrogen has a lower explosive limit (LEL) of 4%. The LEL is defined as the airborne concentration required to support ignition in the presence of sufficient oxygen. Initial DVS data indicated 47 drums had hydrogen concentrations exceeding 4% prior to being purged with nitrogen. Figure 1 presents the projected percentage of drums for which hydrogen levels are expected to exceed the LEL, based on DVS hydrogen data. Waste streams with the highest levels of hydrogen contain cemented inorganics and spent samples (55-38), vacuum filter cake (50-19), and combustible debris (55-19 and 55-30).
Once drums are vented, a hand-held hydrogen monitor is used to monitor hydrogen concentrations through drum vents. The hydrogen monitor has an upper detection limit of 2000 ppm. If concentrations exceed 2000 ppm, a combustible gas indicator is held to the vent to ensure an explosion hazard does not exist. To date, all hydrogen concentrations measured at drum vents have been below 10% of the LEL. All drums are handled from retrieval to venting as if they may contain elevated levels of hydrogen. For example, if the ring that keeps the drum lid attached to the drum is loose prior to the drum being retrieved, as many drum rings are, sparkproof tools are used to tighten the drum ring.
Numerous control measures have been implemented at the TWISP to address hazards at the work site and to ensure maximum protection of workers and the environment. Whenever possible, engineering controls are used first, followed by administrative controls. Personal protective equipment (PPE) is used to supplement engineering and administrative controls, or when such controls are not feasible. Six control measures are presented.
The first control measure, training and communication of job hazards, is required for all personnel involved with TWISP operations. Training details every operational phase and describes the identified hazards for each operation through task hazard analyses. An essential component of personnel training is a description of operational hold points and required actions when hold points are reached. Hold points are defined as unanticipated situations or unusual conditions that arise during operations. When a hold point is reached, work is halted until the TWISP management determines what actions need to be taken before work can safely continue. Examples of hold points include visible liquids from drums or FRP crates, unusual odors, and cracked drum rings. Newly identified operational hazards and changing work conditions are discussed at daily pre-job briefings.
The second control measure is the use of local exhaust and dome ventilation. Local exhaust ventilation allows the removal of hazardous components as close as possible to the source of generation, providing the greatest level of protection. By effectively removing a potential hazard at its source, exposure to personnel is minimized. Portable HEPA filter local exhaust units are placed wherever ES&H personnel determine they are needed. Dome ventilation in the form of a large, stationary HEPA filtration system protects against environmental releases and prevents exposures to the public.
The third control measure consists of engineering controls built into the DVS to minimize hazards during drum venting. Nitrogen purging of drum headspaces containing hydrogen concentrations exceeding the LEL minimizes the possibility of a drum burn during the critical operation of drum venting. Should a drum burn occur, the ASME approved containment vessel provides protection for DVS personnel. In addition, HEPA filtration during drum venting operations minimizes potential particulate and radiological contamination to personnel and the environment.
The fourth control measure consists of identifying, tagging, and prioritizing or segregating "special case" drums. Drums with a suspected or confirmed hazard are identified as "special case". Each "special case" hazard category is represented by a different colored tag that describes the actual or suspected hazard. The colored tagging system permits recognition of the potential hazards from a distance. The tags may also indicate that some type of special action is required. Examples of "special case" drums include suspect pressurized drums, which are scheduled for priority venting, and drums that emit elevated levels of VOCs, which are segregated for more extensive monitoring.
Daily industrial hygiene monitoring is utilized as the fifth control measure. Monitoring is considered a control measure because it provides the means to immediately halt or alter operations where hazardous compounds are suspected or confirmed to be in excess of established action limits. Routine monitoring for VOCs, combustible gases, hydrogen, and carbon monoxide identifies operations where controls are adequate or need improvement.
The use of personal protective equipment (PPE) is the last control measure. PPE is specifically selected for anticipated or identified hazards during each phase of operations. Anti-contamination clothing, respiratory protection, booties, and a variety of gloves are used, depending on anticipated or identified chemical, physical, and/or radiological hazards. Consideration of heat and cold stress is necessary so that, in reducing the risk of one hazard, the risk of another hazard is not elevated to an unacceptable level.
The hazards identified above, combined with the inherent risks encountered at radioactive and hazardous waste disposal sites present unique challenges to ES&H professionals and hazardous waste managers. Based on the initial months of the TWISP operations, the following conclusions are presented to better protect personnel and the environment during drum recovery operations. 1) The likelihood of unvented drums becoming pressurized increases when ambient temperatures increase. 2) Pressurized drums must be vented before they become bulging drums. 3) Vented drums present the potential for VOC emissions and personnel exposure. 4) Large numbers of co-located vented drums may present the potential for increased hydrogen and VOC concentrations within unventilated storage domes. 5) Monitoring and sampling within domes where vented drums are stored is necessary to ensure that the levels of risk to drum handlers and inspection personnel are acceptable. 6) Identifying, tagging, and prioritizing or segregating "special case" drums is necessary to prevent unacceptable personnel exposures and preclude environmental contamination. 7) Acoustic drum pressure detection may be a viable tool in assessing elevated drum pressures.