J.E. Elmer
IT Corporation*

J.E. Virgona (DOE Retired) and M. K. Tucker
U.S. Department of Energy
Grand Junction Office
Grand Junction, Colorado

The U. S. Department of Energy Grand Junction Office has completed the remediation of more than 4,000 properties in the Grand Junction, Colorado, area. Valuable lessons learned during the 15 years of remediation include treatment of hazardous waste, mitigation of radon levels in structures, and dispute resolution. The Grand Junction Office treated more than 15,000 cubic meters (19,000 cubic yards) of hazardous waste on nine properties. As a result of a regulatory ruling requiring all treatment to take place in a container, unique problems such as how to manage large volumes of debris were encountered and resolved. More than 200 properties had elevated radon levels after remediation. Mitigation efforts used to successfully reduce the levels were activating subfloor vents, insulating and ventilating crawl spaces, and retaking radon daughter concentration measurements in the appropriate rooms. The Grand Junction Office and the Colorado Department of Public Health and Environment disagreed on several interpretations of the cleanup standards. Two of the disputes were resolved by convening panels of experts that used different methods to evaluate and resolve the issues.

In 1981, the U.S. Department of Energy Grand Junction Office (DOE-GJO) was assigned responsibility for the Uranium Mill Tailings Remedial Action (UMTRA) Grand Junction Vicinity Properties Project. This project required planning, characterization, design, and remediation of more than 4,000 properties in the vicinity of Grand Junction, Colorado.

The properties were contaminated with radioactive uranium mill tailings (also termed residual radioactive material) as a result of milling operations conducted for the U.S. Government during the 1950s and 1960s. The mill tailings were used by the community as backfill in construction projects, as a sand mixture in concrete and mortar, and as a sandy loam admixture in the native soils to enhance the growth of gardens and lawns. Congress passed the Uranium Mill Tailings Radiation Control Act (UMTRCA) [1] in 1978 to authorize DOE to remediate and to permanently dispose of the radioactive mill tailings that were spread primarily throughout the southwestern United States.

DOE-GJO and its contractors were given the task of planning and performing the cleanup of more than 4,000 residences and businesses, the largest remedial action project of its kind. After 15 years of remediating properties, several of the lessons learned include treatment of hazardous waste on 9 properties, mitigation efforts to reduce elevated radon levels in more than 200 structures, and resolution of major issues with the Colorado Department of Public Health and Environment (CDPHE).

During the project, hazardous waste deposits commingled with uranium mill tailings (commingled waste) were discovered while vicinity properties were being surveyed and remediated. Only 20 deposits of commingled waste were identified on the 4,301 properties surveyed. The deposits can be summarized as follows:

• 9 properties contained deposits of Resource Conservation and Recovery Act (RCRA) characteristic wastes (lead, chromium, cadmium, and arsenic)
• 4 deposits of potential RCRA-listed wastes; process knowledge was used to determine that these wastes were not hazardous waste
• 2 properties contained polychlorinated biphenyls (PCBs) above Toxic Substance Control Act (TSCA)-regulated limits
• 2 properties contained solvents (RCRA-listed wastes)
• 1 property contained characteristic wastes treated by the U.S. Environmental Protection Agency (EPA) under a Superfund emergency action
• 1 property contained characteristic hazardous waste that is being treated by DOE in a treatability study
• 1 property contained listed hazardous waste that was successfully treated in a treatability study

All UMTRA disposal cells are designed to meet the U.S. Nuclear Regulatory Commission (NRC) and EPA requirements for disposal of residual radioactive material (RRM). Although the cells are designed for a 1,000-year design life, they do not meet RCRA standards for a hazardous waste disposal facility. Most of the UMTRA cells, including the Cheney disposal cell near Grand Junction, do not have synthetic or double liners. Consequently, RRM mixed with a hazardous waste had to be treated to a nonhazardous waste standard so that the remaining RRM could be disposed of at the Cheney cell.

DOE made the programmatic decision to only treat RCRA characteristic hazardous wastes because it was technically practical to treat the waste at a relatively low cost. RCRA-listed wastes and PCBs were not treated because of the high cost and low probability of delisting a listed waste and because of the difficulty in disposing of PCB wastes. To treat characteristic hazardous waste, CDPHE initially required a consent agreement between the two agencies in lieu of a RCRA permit. Both parties sought to ensure that their respective liabilities were minimized. DOE was not willing to enter into more risk than it believed necessary to treat the hazardous waste because DOE was not the generator of the waste. CDPHE was not willing to relinquish its regulatory oversight control and reserved the right to pursue enforcement action if DOE did not perform to the requirements of the agreement.

In 1996, CDPHE made the regulatory interpretation that DOE could perform hazardous waste treatment without a permit or consent agreement under the provisions of Permit by Rule (6 Colorado Code of Regulations 1007-3 Part 100.21). Permit by Rule required DOE to submit a Waste Analysis Plan to comply with Part 262 of the Colorado Hazardous Waste Act and to perform all treatment in an accumulation tank or container. Because of the container requirement, CDPHE did not allow use of a pug mill that would have increased production rates with a more efficient system. Because of on-site storage requirements, commingled waste material could not be transported to nearby properties for treatment at a centralized location. This restriction resulted in construction of separate decontamination pads and laydown areas at each property and in mobilization of treatment equipment to each site.

The general process for hazardous waste treatment at each site involved the following steps:

• Excavate commingled waste with appropriate dust controls
• Run material through screens to isolate debris and rubble exceeding 6.4 centimeters (2.5 inches) in diameter
• Convey 6.4 centimeter (2.5-inch) minus material to mixer
• Add reagent (typically Portland cement) and water to mixer
• Operate mixer for 5 to 10 minutes to ensure thorough mixing
• Unload treated material into stockpiles
• Treat debris with high-pressure wash system
• Sample and conduct laboratory analyses to demonstrate material was successfully treated

Table I presents summaries of the volumes treated, original concentration of hazardous waste, and final treatment results for nine properties. The project successfully treated more than 15,000 cubic meters (19,700 cubic yards) of hazardous waste mixed with RRM. The lessons learned can be categorized into two different areas: treatment and debris.

Because of the requirement to use a container for treatment, truck-mounted commercial cement mixers were chosen to perform the work on the larger properties. Cement mixers are designed to mix and discharge a slurry; however, the objective of the treatment was to stabilize the materials through chemical bonds by mixing the materials until they formed a loose solid. The materials being treated were essentially silty clays or sandy silts mixed with tailings. The treatment process specified low quantities of water, up to 10-percent moisture content, so that the end product was typically stiff (no slump). Problems with materials sticking to the interior of the drums and long unloading times of the treated material were commonplace.

When the subcontractors did not monitor the mixing process in the drum, the treated mixture would build up on the inside until the fins of the mixer were no longer effective. It then typically took several days for laborers to chip out the solidified material stuck to the drum. Several subcontractors used form oil on the drum interior to help prevent the material from sticking. Also, the size and angles of the mixer truck's fins affected the unloading times. The fins with the greatest slope discharged the best.

The process was sensitive to the moisture content of the soils. Wetting the soil sufficiently before loading the material into the mixer helped reduce buildup in the mixer. If water was added to the mixer after the material was added, it would form a cement slurry that bound to the mixer and not the material. Because some excavated materials had a higher moisture content, it was imperative to take daily moisture tests of the soils to determine how much water had to be added to the mixture. Ideally, enough water was needed so that the end mixture formed small clumps. Too much water caused the slurry to stick to the walls; too little water and it did not hydrate the cement. On most properties, the ideal moisture content was less than 10 percent. At the river embankment property, moisture content had to be increased to 15 percent because of the clayey nature of the soils.

Several properties located along the riverfront contained large amounts of buried trash and debris. One property was used as a general dumping ground for household trash, while another property contained large quantities of construction debris (e.g., concrete, asphalt). Debris generated from both properties required a considerable amount of special handling and decontaminating. The large quantity of debris slowed down production because it would constantly plug the vibrating screens used to separate debris from soil.

Lined pads had to be set up to contain the wash-down water and sludge washed off the debris because both mediums were considered hazardous waste. One subcontractor designed an innovative bucket to act as a screen so that debris could be picked up, washed while in the bucket, and then placed in a stockpile. With this method, the debris was only handled once.

More than 20 percent of the material excavated, approximately 230 cubic meters (300 cubic yards), on the riverfront property was larger than 6.4 centimeters (2.5 inches) in diameter and had to be managed as debris. This material included river cobble, metal, wood, asphalt, concrete, and plastic battery casings. A biased sampling regime was established to sample only the plastic battery casings. Although the Toxicity Characteristic Leaching Procedure (TCLP) value of lead for the battery casings was 9.6 milligrams per liter (mg/L), the average for this waste stream was well below the regulatory threshold of 5.0 mg/L for lead. Because the debris had been radiologically decontaminated, the material was accepted by the local landfill for disposal.

More than 14 percent (approximately 100 cubic meters [130 cubic yards]) of the material excavated at the insurance yard property was debris. Battery pieces were also sampled from the waste stream at this site. The TCLP value of lead for this waste stream was 0.6 mg/L, which was well below the standard. Because this material had been radiologically decontaminated, it could be disposed of at the local landfill.

The EPA standard for UMTRA, 40 CFR 192 [2], states that "the objective of remedial action shall be, and reasonable effort shall be made to achieve, an annual average radon decay product concentration not to exceed 0.02 working level (WL). In any case, the radon decay product concentration shall not exceed 0.03 WL." The standard also states that RRM should be removed from buildings exceeding 0.03 WL; however, sealants, filtration, and ventilation devices may provide reasonable assurance of reductions from 0.03 WL to below 0.02 WL. EPA allowed the indoor levels to exceed the standards when the sources were other than RRM (i.e., naturally occurring).

Every property that underwent remediation had final radon daughter concentration (RDC) measurements taken in the habitable structures located on the property. Of 4,190 properties measured, 203 had elevated RDC measurements. Approximately 142 of the properties had RDC measurements between 0.02 and 0.03 WL. Only 61 properties (less than 2 percent of the total number of properties measured) had elevated RDCs above 0.03 WL after remediation was performed. Because most properties were included into the UMTRA project as a result of exceeding the exterior radium-in-soil standard or interior gamma standard, the project did not develop a large quantity of data documenting RDC conditions before remediation. Thus, it was difficult to assess if these properties had a preexisting RDC problem.

DOE's interpretation of the standard was that if RRM was removed and the RDC level was decreased to less than 0.03 WL, the standard had been met. CDPHE contested that the standard of 0.02 WL had to be met and that additional mitigation would have to be performed. DOE and CDPHE performed a paper review of 200 properties with RDC levels between 0.02 and 0.03 WL. By reviewing the radiological assessments, designs, and remedial action, the agencies were able to identify several alternatives to potentially lower the RDC. After several years of negotiation, DOE agreed to perform "simple fixes" on properties to attempt to lower the RDC to below the 0.02 WL. Simple fixes were defined as low-cost measures, such as

• Insulating and ventilating crawl spaces, including isolating crawl spaces from basements by sealing openings such as trap doors, pipe penetrations, or gaps in walls and ceilings.
• Activating subfloor radon vent systems previously installed but not activated (typically just capped at the floor).
• Reassessing the area within 15 meters (50 feet) of the structure for gamma anomalies that might indicate deposits that were missed; RRM above the standard would then be removed.
• Retaking the RDC measurement in a different location if the original placement of the measurement device was in a room without ventilation or not considered habitable.
• Identifying other sources, such as a rock collection containing uranium ore, that would affect the RDC levels.

Although the paper study identified appropriate alternatives, there was typically insufficient documentation to determine if the structure included a crawl space that could be ventilated. Consequently, a field visit was required to determine what steps could be taken. Figure 2 presents the logic for making the decision on which simple fix to perform.

Generally, simple fixes were successful in reducing radon levels. The following list summarizes the results of implementing simple fixes to reduce RDC levels in structures:

Most crawl spaces inspected were accessible with 0.6 to 0.9 meter (2 to 3 feet) of clearance for workers to enter and install the insulation under the floor boards of the structure. Crawl spaces with less than 0.6 meter (2 feet) of clearance were not practical to insulate and did not qualify for a simple fix. Vents were normally added so that there would be at least one per side of the structure. In some cases, existing crawl-space vents could be retrofitted, but most vents had to be installed by cutting through the concrete foundation. Because of the size of crawl spaces, costs varied dramatically from $2,000 to $10,000. Insulating and ventilating the crawl spaces dropped the average RDC value in the structures by more than 39 percent. Nineteen of the 21 structures measured had final RDC values below the 0.02 WL standard. Deleting the two RDC values that increased, the average RDC in the remaining 19 structures dropped by an average of 56 percent.

By far the easiest mitigation effort was to activate the vent system that was installed during remediation of the interior. Vent systems were considered cheap insurance and were installed during reconstruction of all interior concrete slabs. Because many owners did not want the vent risers to extend through their roofs, the vents were not activated unless they were needed. Generally, removing the RRM source was sufficient to lower the radon levels. Activating the vents lowered the RDC level in the structure by more than 70 percent, even though the vent may not have been installed under the entire house. (In some structures, tailings were used only under an addition.) RDC levels were lowered below the 0.02 WL standard in 27 out of 29 properties that were measured.

Properties with no apparent simple fix, such as activating a vent or insulating and ventilating a crawl space, were reassessed for radiological contamination to ensure that reasonable steps had been taken to identify all sources of RRM. Because of the potential for soil gases to migrate to a structure, the area within 15 meters (50 feet) of the structure was studied. The reassessment protocol called for hand excavating over suspect utilities, drilling to the bottom of the foundation on each side of the house, rescanning the interior of the house, and investigating gamma anomalies that might indicate a hidden deposit. Usually, only small deposits were detected. Of the 25 properties with deposits removed, 11 of the properties had final RDC levels of more than 0.02 WL, 5 of which were more than 0.03 WL. In the data set, 5 of the 25 RDCs increased, indicating the action either had no effect and there was a substantial naturally occurring source in the soils or there was an error in the measurement. The average reduction in the RDC after removal of the deposits was 31 percent.

During the review of the properties with elevated readings, it became apparent that many of the measurement devices were placed in rooms that were not indicative of the average habitable condition of the house. The programmatic guidelines specified that the device should be placed in the lowest habitable space. Sometimes the devices were placed in basements that were not finished and often used only for storage. Since most of the structures were 30 to 40 years old, these basement rooms often had no ventilation or windows and usually had only one access. These rooms often had openings to exposed dirt that would allow radon from naturally occurring sources to enter the structure. Measuring the RDC for the structure in such rooms would not be representative of an annual average exposure to an individual. New RDC measurements were taken in 22 structures because the original readings were obtained in rooms later defined as not habitable. The new readings resulted in more than a 50 percent drop in the radon measurement for these structures. Twenty-one of the 22 new measurements were below the 0.02 WL standard.

Many properties contained rock collections with radioactive ores. In most cases the property owner was convinced to dispose of the rocks during remedial action. Several rock collections were discovered in structures with elevated RDCs. On two properties, the owners allowed the rock collections to be removed. The resulting RDC dropped an average of 43 percent.

One structure presents an interesting case study on the effects of remediation and where to measure RDCs (see Table II). The pre-remedial action average RDC for the structure was 0.1549 WL. The highest reading in the structure, 0.1720 WL, was located in the bedroom (Room J) where the highest gamma signature in the structure was measured. Remedial action removed significant exterior and interior deposits. A large interior deposit was removed in the south portion of the house; Room J was located at the north end of the house and a small deposit was removed from this room. After the initial remedial action, an RDC of 0.0597 WL was measured in Room J. Because this reading exceeded the 0.03-WL action level, the vent system previously installed in the south portion of the house was activated. The new reading in Room J, taken after the vent activation, was 0.165 WL. An interview with the owner of the house revealed the bedroom was not in use and the door had been sealed, completely isolating the room and preventing any movement of air in the room. Thus, the measurement did not reflect the average habitable condition of the house. To ensure that the next measurements were correct, Room J was opened up and six monitors were placed around the house to obtain an average. The new average was 0.0131 WL, well below the standard. The measurement in Room J decreased tenfold to 0.0105 WL, demonstrating the effects of ventilation and the necessity of taking a measurement in the proper room.

DOE and CDPHE disagreed on several interpretations of the EPA cleanup standards, 40 CFR 192. UMTRCA legislation required the State to pay 10 percent of remedial action costs, while specifying that the State shall participate fully in the selection and implementation of remedial action. This requirement gave the State a unique role of being regulator and partner in the program. Because the State signed all remedial action agreements with property owners, the State did not have to sign the agreement if it disagreed with the cleanup strategy. Thus, the State had the ability to affect decision making at the individual property level.

Like most environmental restoration projects, the argument of how clean is clean often arose. The State, pursuing the best interest of public health, pushed for lower standards and a more conservative interpretation of the regulations. Because DOE and CDPHE agreed to disagree on several issues during the 15-year life of the program, several techniques were used to resolve the issues. The simpler techniques were to hold quarterly manager meetings and convene a technical liaison committee to resolve issues. Another technique, described in more detail, was to convene panels of experts.

An issue panel was established to resolve two outstanding technical issues. The panel was composed of a technical representative from each agency. To qualify as a technical representative, each person had to have a radiological background but could not work directly in the UMTRA Program. Both DOE and CDPHE prepared issue papers that described their respective positions. The parties then presented their arguments orally to the panel. A final report [3] was issued by the panel documenting its decision. Both parties agreed to the findings, resolving two long-standing issues.

The first issue involved the use of delta scintillometer measurements on 12 properties where DOE determined there was no contamination exceeding EPA standards. Thus, the properties were excluded from remediation. CDPHE argued that the delta measurement was inappropriate to use in deciding whether to include or exclude a property. Soil samples were considered the best measurement but were not always used early in the program. DOE argued that the delta value could be a reliable measurement by using a conservative correction factor.

The panel determined that the delta measurement was consistent with the guidance of 40 CFR 192, which states that gamma measurements can be utilized above and below the ground to demonstrate compliance. The panel agreed that the 12 properties being contested should stand as no actions.

The second issue involved radiological assessments and excavation control methods. CDPHE believed that DOE's assessments did not target enough borings in areas that were likely places of fill and did not investigate enough of the gamma anomalies measured in excavations to determine if hidden deposits still existed. CDPHE also cited that DOE's two remedial action contractors differed in their excavation practices and that they should be more consistent. DOE argued that its assessment and excavation practices met the intent of the EPA standard, which required reasonableness in the approaches and does not require 100 percent assurance that all tailings were found.

The panel determined that DOE should implement one technical procedure for excavation control practices. It also found that DOE and CDPHE should sign a formal agreement if they wished to establish additional requirements above 40 CFR 192.

As a result of the decisions made by the issue panel, DOE did not have to remediate 12 additional properties, resulting in a cost avoidance of $330,000. DOE did agree to discontinue using the delta scintillometer measurements to make inclusion/exclusion decisions. However, DOE did not change its excavation control practices because the panel determined that both contractors' practices met the requirements of 40 CFR 192. DOE did alter assessment practices slightly to target more borings in areas of fill, such as around foundations and utilities.

In 1991, an environmental audit raised questions concerning the adequacy of the UMTRA Program vicinity properties inclusion/exclusion process. A Peer Review Panel of three outside experts was contracted by the DOE to review the process and determine whether there was reasonable assurance that properties with contamination in excess of the standards were being included in the UMTRA Program and remediated. The panel reviewed how soil samples were being analyzed, the use of gamma measurements for finding buried deposits, the use of RDC measurements for determining inclusion/exclusion status of a property, and the likelihood of missing shielded deposits.

The final report [4] issued by the panel documented that although field soil sample analysis results did not have a high correlation to laboratory data, the process almost always overestimated the true concentration, so reasonable assurance was still provided that properties were not falsely excluded. The panel also determined that the gamma screening techniques provided reasonable assurance that large hidden deposits were being identified. The panel, however, did observe that a property could be excluded from remediation without taking an RDC measurement. Thus, reasonable assurance was not provided to ensure that an elevated radon level could exist from a deposit that was shielded from assessment. The panel recommended that a sample of excluded properties be measured to determine if structures had been falsely excluded. After debating the issue with CDPHE, DOE agreed to perform radon sampling in excluded structures to determine if a significant number had been falsely excluded.

The size of the sampling population and criteria for determining if too many properties were falsely excluded was heavily debated between CDPHE and DOE. DOE targeted a group of 233 properties based on high inside gamma readings (greater than 15 microroentgens per hour), expecting a possible correlation between the high readings and radon. Although DOE attempted to choose 100 properties randomly out of the group to measure, all 233 property owners were contacted before 94 property owners allowed access to their houses. Of the 94 houses measured, 9 of the structures had resulting RDC values exceeding 0.02 WL. The agreement was then reached that if less than 5 percent of the structures (equates to five structures) were assessed and found to have RRM exceeding standards, no more sampling would have to be conducted. If 5 percent or more had RRM that exceeded the standard, DOE would have to re-sample the remaining 139 properties, plus potentially revisit all excluded structures.

The potential affect on the program was enormous if the results showed that 4,000 excluded properties had to be re-sampled for radon. A statistical correlation between gamma measurements and radon detected in structures suggested up to 5 percent (or 277) of the properties could exceed the 0.02-WL standard and would need to be included in the project for remediation. Because all 4,000 properties could not be sampled at once, an estimated 3 to 4 years would be required to re-sample, design, and remediate the potential 277 properties at a cost of $15,000,000.

Of the nine properties that had to be reassessed, only six had RRM that exceeded standards. Of the six properties, only one had a significant deposit (38 cubic meters [50 cubic yards]). After remediation of the six properties, RDC levels for only three of the structures fell below 0.02 WL. The other three RDCs increased to levels above 0.04 WL, demonstrating that the use of radon measurements was not an accurate tool for demonstrating the presence of RRM. All the buried contamination found on the six properties was in shallow deposits of 30 centimeters (12 inches) or less. Thus, we can conclude from the small statistical sample, that hidden deposits were not contributing to the elevated radon levels in the structures. The contamination was either at such low levels or in small deposits that it was probably missed during the original surveys because of field screening techniques used at the time.

Because the RDC measurements in nine of the structures exceeded the criterion of five, DOE was obligated to re-sample the remaining population. Of the remaining 139 properties, the DOE contractor was able to gain access to only 13 properties to place measurement devices. None of the final results from the 13 properties exceeded 0.02 WL.

CDPHE agreed that the results were not significant enough to require additional sampling on the remaining properties, closing what could have been a major issue that could have extended the end of the UMTRA Program. Six years evolved since the original environmental audit to close out the issue, which resulted in the remediation of six additional properties, less than 0.2 percent of the total properties remediated to date.

The Grand Junction Vicinity Properties Project is coming to a successful close in 1998 after remediating more than 4,000 properties. Lessons learned during the 15-year program can be applied to other environmental restoration programs. Because of the unique regulatory determination made by CDPHE, DOE had to treat 15,000 cubic meters of hazardous waste in large mixers. Experience showed that special attention had to be paid to the moisture content of the material. Additional requirements were also necessary to manage the large quantities of debris generated so that it could be disposed of at a licensed landfill. DOE agreed with CDPHE to attempt to lower elevated radon levels in more than 200 structures. Effective mitigation efforts included activating subfloor vent systems, insulating and ventilating crawl spaces, retaking RDC measurements in the proper rooms, and reassessing and removing additional deposits of contamination. To resolve several outstanding issues with CDPHE on interpretation of the cleanup standards, DOE convened two different panels of experts. These panels were able to resolve the issues that had the potential to affect the closure of the program.

1. 42 U.S.C., Uranium Mill Tailings Control Act of 1978, Public Law 95-604, as amended, Section 7901, et seq., and Section 7942, United States Code.
2. 40 CFR 192, U.S. Environmental Protection Agency, "Health and Environmental Protection Standards for Uranium and Thorium Mill Tailings," U.S. Code of Federal Regulations.
3. T.B. BORAK, D.L. DUNCAN, and N. SAVIGNAC, UMTRA Project Vicinity Property Inclusion/Exclusion Process, Final Report, Peer Review Panel, prepared for Jacobs Engineering, Albuquerque, New Mexico (January 8, 1993).
4. G.E. RUNKLE and D.H. SIMPSON, Final Report of the Panel to Resolve Two Issues Between the Colorado Department of Public Health and Environment and the United States Department of Energy, U.S. Department of Energy (G.E. Runkle) and Colorado Department of Public Health and Environment (D.H. Simpson) (October 17, 1994).

BACK