WM’00 Conference, February 27-March 2, 2000, Tucson, AZ

FREEZE/THAW IMPACTS ON RADON DIFFUSION CHARACTERISTICS OF COVER SOIL ON THE UMTRA PROJECT Douglas E. Gonzales Jacobs Engineering Group Inc. 125 Broadway Ave., Oak Ridge, Tennessee 37830, Jere B. Millard Terradigm, Inc. 401 Alvarado, S.E., Albuquerque, New Mexico 87108, and Ronald E. Rager Morrison Knudson Corporation Morrison Knudson Plaza, Boise, Idaho 83729

ABSTRACT A simple study was performed to access the effects of freeze/thaw cycles on radon diffusion coefficients for unsaturated cover soil used to environmentally isolate uranium mill tailings on the Uranium Mill Tailings Remedial Action (UMTRA) Project. Typical radon barrier soil (Estes Gultch barrow source, Rifle, CO) was compacted to 95-percent Proctor in a 4-inch high, 4-inch diameter cylindrical steel mold used to perform standard transient radon diffusion coefficient measurements. Samples were prepared at 4.4, 9.4, and 13.7 percent by weight moisture content, corresponding to 22, 46, and 68 percent moisture saturation, respectively. The compacted, moisture conditioned soil samples were sealed in the molds by spring-loaded steel plates, and the samples were subjected to 13 rapid freeze/thaw cycles using the freezer compartment of a conventional, domestic refrigerator: 15 hours freezing followed by 9 hours thawing. In the process, soil temperatures cycled between -12 and 20 degrees centigrade. After 13 cycles, the diffusion coefficient for each sample was measured using the transient diffusion method and compared to corresponding measurements made prior to initiating the cyclical freeze/thaw treatment. The results of this freeze/thaw experiment indicate that the freeze/thaw cycling of typical radon barrier soil, with moisture saturation less than 75 percent, would not significantly alter the radon diffusion properties of the cover soil. The relationship of these findings to observations noted by others and their application to the UMTRA Project is discussed. INTRODUCTION The design of disposal facilities for the long-term environmental isolation of uranium mill tailings, contaminated soil and mine wastes, or any other wastes containing radium226 incorporate multi-layered covers to fulfill certain performance objectives. Performance objectives are established to protect the waste from being dispersed by natural forces, prevent intrusion of plants and animals into the waste, control infiltration

WM’00 Conference, February 27-March 2, 2000, Tucson, AZ

of water through the encapsulated waste, inhibit waste constituents from leaching into groundwater, and limit the surface flux of radon-222 into the atmosphere. The waste disposal facility cover is the main engineered component responsible for attaining these objectives. The engineered, multi-layered cover is typically constructed of different natural and synthetic materials that collectively support the attainment of the stated performance objectives. Functionally, the cover components above the lowest lateral drain/filter layer serve to protect the natural erosion of the waste by promoting controlled vegetative growth on the cover surface and prevent biointrusion of the plant roots into the waste as well as the intrusion of animals (including humans). More importantly, these cover components protect the integrity of the primary water infiltration/radon diffusion barrier overlying the buried waste. The filter/drain layer sandwiched between the erosion protection/rooting media/biointrusion layers and the infiltration/radon barrier is designed to reduce the hydraulic head on the infiltration barrier, and, therefore, reduce the infiltration of water through the tailings. A composite cover containing representative design componentss is shown in Figure 1. Vegetation 0.5’ - 1.0’

Erosion Barrier:

Gravel*

0.5’ - 2.0’

Root Zone:

Selected Soil*

0.5’

Filter/Drain/Bedding:

Sand

0.5’ - 4.0’

Frost Protection/ Secondary Root Zone:

Random Soil

0.5’

Filter/Drain/Bedding:

Sand

1.0’ - 3.0’

Drain/Biobarrier:

Gravel & Cobbles

1.0’

Filter/Drain/Bedding: Infiltration Barrier:

Sand Geomembrane, or Geomat, or Silt & Clay

1.5’ - 2.0’

Radon/Infiltration Barrier:

Silt & Clay

Waste:

* In some instances, these two components may be inverted or mixed

Figure 1.

Composite cover containing representative design components.

To attain a low hydraulic conductivity (90 percent. Therefore, based on analyses of freeze/thaw behavior of clayey soil, the effect of freeze/thaw on permeability; and the lack of any testing to determine the effect on the radon flux rate following freeze/thaw of radon barriers; the following are considered when designing and constructing covers in locations where freeze/thaw cycling is likely: 1.

Determine the frost penetration depth through modeling (4), and place sufficient soil above the infiltration/radon barrier to accommodate the modeled frost penetration depth, or

2.

Design the cover thickness for radon attenuation through modeling (7) based on 90% compaction rather than the conventional 95% (Proctor) compaction to simulate a potential increase in effective porosity due to soil that has been expanded by freeze/thaw cycles.

If the infiltration/radon barrier is designed to be located below the frost depth (as per 1 above), using a reduced compaction to ascertain the thickness the barrier through radon diffusion modeling, is not necessary. However, at sites located in the arid western United States, infiltration/radon barriers are not saturated following construction, nor have limited field measurements indicated that they exhibit long-term saturation. The effects of freeze/thaw on infiltration/radon barriers and required protection for unsaturated conditions that actually exist are not well understood. Permeability testing following freezing and thawing of soils in the range of 60-70% of saturation, as expected to exist at these sites has not been performed.

WM’00 Conference, February 27-March 2, 2000, Tucson, AZ

Also, field permeability tests using double ring infiltrometers on test fills only partially frozen and thawed in depth, indicate that the hydraulic conductivity is basically unaffected by the frozen upper portion of the barrier soils (5). This is because the overall hydraulic conductivity of the barrier is controlled by the lowest permeable element. This leads one to conclude that in order to effectively limit infiltration, the entire thickness of the radon/infiltration barrier need not be protected from freezing. Consequently, designing covers containing combined infiltration/radon barriers based on assumed freeze/thaw impacts for saturated and near saturated soil, may not be necessary or cost-effective. The following simple testing was performed to assess the effects of freeze/thaw cycles on the radon diffusion coefficients for unsaturated cover soil with varying degrees of saturation. EXPERIMENTAL DESIGN AND RESULTS Radon diffusion measurements were made for typical radon barrier soil used on the UMTRA Project, which had, and had not been subject to a series of rapid freeze/thaw cycles. Radon barrier soil from the Estes Gulch barrow source, Rifle, CO was compacted to 95-percent Proctor in 4-inch high, 4-inch diameter cylindrical steel molds used to performing standard transient radon diffusion coefficient measurements. Samples were prepared at 4.4, 9.4, and 13.7 percent moisture content by weight using a conventional microwave oven, and correspond to 22, 46, and 68 percent moisture saturation, respectively. The physical properties and radon diffusions for radon barrier soil before and after freeze/thaw treatment are presented in Table 1. The compacted, moisture conditioned soil samples were sealed into the radon diffusion coefficient testing molds by spring-loaded steel plates to simulate soil compression that might occur if a radon barrier were placed beneath other cover components used for erosion protection, plant rooting media, and biointrusion. The complete, assembled test module is shown at the left in Figure 2, where the cylindrical mold used in the transient radon diffusion measurements is at the right.

WM’00 Conference, February 27-March 2, 2000, Tucson, AZ

Table I

Physical and radon diffusion properties of radon barrier soil before and after freeze/thaw treatment (specific gravity = 2.68 g/cm3 ).

Moisture Content Density Porosity Saturation Diffusion Coefficient (% dry weight) (g/cm3 ) (%) (cm2 /sec.) -----------------------------------------------------------------------------------------------------------Before Treatment 6.0

1.73

0.35

29

0.028

9.5

1.74

0.35

47

0.016

10.9

1.75

0.35

55

0.019

15.3

1.73

0.35

75

0.0012

4.4

1.74

0.35

22

0.028

9.4

1.73

0.35

46

0.018

13.7

1.74

0.35

68

0.0032

After Treatment

Figure 2.

Complete test assembly for freeze/thaw assessment of radon diffusion coefficient variability as a function of moisture saturation.

WM’00 Conference, February 27-March 2, 2000, Tucson, AZ

The samples were placed in the freezer compartment of a conventional refrigerator and subjected to 13 rapid freeze/thaw cycles: 15 hours freezing followed by 9 hours thawing. In the process, the average soil temperatures cycled between -12 and 20 degrees centigrade. After 13 cycles, the radon diffusion coefficient for each sample was measured and compared to corresponding measurements made prior to initiating the cyclical freeze/thaw treatment. The transient radon diffusion method (8) was used to measure the radon diffusion coefficient. This method consists of (1) compacting a sample of soil into a test module (the 4” high, 4” diameter steel cylinder shown at the left in Figure 2), (2) placing the module on top of a known radon source, and (3) measuring the time-dependent diffusion of radon into a fixed volume above the compacted soil sample to determine the diffusion coefficient at the prepared compaction and moisture saturation. Figure 3 is a schematic of the diffusion coefficient measurement apparatus.

Figure 3.

Schematic of the transient radon diffusion apparatus (8).

WM’00 Conference, February 27-March 2, 2000, Tucson, AZ

The results of the radon diffusion coefficient measurements for radon barrier soil subjected to 13 rapid freeze/thaw cycles are given in Table 1, as a function of percent moisture saturation. Previously performed radon diffusion coefficient measurements made on soil samples from the same borrow source that had not been subjected to freeze/thaw cycling are also given in the Before Treatment section of Table 1. Figure 4 compares the Before and After Treatment radon diffusion coefficient measurements as a function of fractional moisture saturation. INTERPRETATION OF RESULTS AND CONCLUSIONS Comparing radon diffusion measurements for soil that has been subjected to freeze/thaw treated and untreated soil demonstrated that the effect of freeze/thaw cycles on radon diffusion characteristics of unsaturated radon barrier soil is minimal for moisture saturation less than 75 percent saturation. Estimated long term predicted soil moisture contents and results of this test program, indicate that once equilibrium moisture is established in the radon barrier, typically less than 75 percent saturation, there will be little or no effect on radon flux rate. However, since initial placement conditions result in wetter soil conditions (greater than 80% saturation), it is still prudent to account for the effect of freezing. Although the freeze/thaw tests performed were limited and simple, they provide assurance that the UMTRA designs based on placing the infiltration/radon barrier below the modeled frost penetration depth and assuming a 5 – 10 percent increase in porosity due to freeze induced soil expansion are conservative. Neither the rate of freezing nor direction of freezing was controlled in these simple experiments. Freezing may have been accelerated and variable in the refrigerator. More rapid rate of freezing should lead to maximum disturbance of the soil sample due to stresses induced at the freezing front and from the formation of ice lenses. However, this was not the case in the simple tests performed. The porosity of the samples, an indication of disturbance, remained constant after being subjected to freeze/thaw cycles, indicating an absence of sample disturbance. It appears that the porosity of the soil is an indicator of how freezing and thawing disturbs the soil in the laboratory tests. As previously noted, experience of freezing saturated soils (DOE, 1988) can result in an increase in porosity. Additional investigation of the effects of freeze/thaw is justified to reduce or eliminate the use of additional cover soil used solely for the frost protection of combined radon/infiltration barriers under unsaturated conditions. These should include additional testing to extend the results to near saturated and saturated conditions, as well as additional control of the freezing rate and direction. At the Weldon Spring CERCLA site, the design of a mixed radioactive-hazardous waste disposal cell includes a compacted clay radon/infiltration barrier. The barrier is three feet thick and is only partially protected from freezing in extreme events by overlying

WM’00 Conference, February 27-March 2, 2000, Tucson, AZ

protective layers. The barrier is placed at 85% saturation and compacted to 95% standard Proctor. Since the site is located in the central U.S., long term moisture content will not

Figure 4.

Comparison of radon diffusion coefficientfs for soil with and without freeze/thaw cycling, as a function of fractional moisture saturation.

WM’00 Conference, February 27-March 2, 2000, Tucson, AZ

vary from initial conditions. Under extreme conditions (estimated at 200-year return event), freezing of the upper 12-inches of the radon/infiltration barrier will occur. This is predicted to have little effect on the long-term infiltration rate of the cover because the unfrozen lower portion of the barrier will control the infiltration rate. Construction of the radon/infiltration barrier is underway, with the first one-toot thickness placed as an interim winter barrier over one-half of the approximate 24 acre barrier. This barrier will be allowed to freeze and will be field-tested for radon flux next spring. While it is planned to reprocess the upper 8-inches of the interim cover, the flux testing will be performed on the frozen and thawed material. While no temperature monitoring is planned, previous site experience indicates that numerous freeze/thaw cycles of this shallow barrier will occur. Therefore control tests are planned where the barrier surface will be protected from freezing using straw bales covered with plastic sheeting. This program will provide an additional body of data on freeze/thaw effects on radon diffusion by extending the tests into the near saturated and in situ field conditions.

Acknowledgements - This work was supported by DOE under Contract No. DE-AC0491AL62350

WM’00 Conference, February 27-March 2, 2000, Tucson, AZ

REFERENCES 1. Daniel, D and Benson, C., 1990. “Water Content-density Criteria for Compacted Soil liners”, Journal of Geotechnical Engineering, Vol. 116, No. 12, American Society of Civil Engineers, Reston, Virginia. 2. U.S. Department of Energy, 1989. “Moisture Content and Unsaturated Conditions in UMTRA Project Radon Barriers,” Technical Assistance Contractor, UMTRA Project Office, Albuquerque, New Mexico. 3. U.S. Department of Energy, 1989. “Technical Approach Document,” UMTRADOE/AL 050425.002, Revision 2, Technical Assistance Contractor, UMTRA Project Office, Albuquerque, New Mexico. 4. U.S. Department of Energy, 1988. “Effect of Freezing and Thawing on UMTRA Covers, A Special Study Finding,” Technical Assistance Contractor, UMTRA Project Office, Albuquerque, New Mexico 5. Benson, C.H., Abichou, T., Olson, M., and Bosscher, P., 1995. “Winter Effects on Hydraulic Conductivity of Compacted Clay”, Journal of Geotechnical Engineering , Vol. 121, No. 1, American Society of Civil Engineers, Reston, Virgina. 6. Lambe, T. and Whitman, R., 1969. “Soil Mechanics”, Massachusetts Institute of Technology, John Wiley & Sons, Inc., New York, New York. 7. U.S. Nuclear Regulatory Commission, 1984. Radon Attenuation Handbook for Uranium Mill Tailings Cover Design, NUREG/CR-3533 prepared by Rogers and Associates Engineering, Salt Lake City, Utah for the U.S. Nuclear Regulatory Commission, Washington, D.C. 8. U.S. Nuclear Regulatory Commission,1982. Comparison of Radon Diffusion Coefficient Measured by Transient Diffusion and Steady-State Laboratory Methods, NUREG/CR-2875, prepared by K.K. Nielson, D.C. Rich, and V.C. Rogers, Rogers and Associates Engineering, Salt Lake City, Utah for the U.S. Nuclear Regulatory Commission, Washington, D.C.