ML19087A171
| ML19087A171 | |
| Person / Time | |
|---|---|
| Site: | PROJ0734 |
| Issue date: | 07/01/2018 |
| From: | Lloyd Desotell Division of Decommissioning, Uranium Recovery and Waste Programs |
| To: | |
| Desotell L | |
| References | |
| Download: ML19087A171 (5) | |
Text
07/2018 NRC Staff Preliminary Comments on DOE Document, SRRA107772-000009, Rev. A p. 1 NRC Staff Preliminary Comments on DOE Document, SRRA107772-000009, Rev. A, "Predicting Long-Term Percolation from the SDF Closure Cap" dated 04/23/2018
Purpose:
During the U.S. Nuclear Regulatory Commission (NRC) July 2018 Onsite Observation Visit (OOV) to the U.S. Department of Energy (DOE) Savannah River Site Saltstone Disposal Facility (SDF), the DOE requested to receive as soon as possible the NRC staff preliminary comments on the DOE document, SRRA107772-000009, Rev. A, Predicting Long-Term Percolation from the SDF Closure Cap, dated April 23, 2018. The DOE document is available in the NRCs Agencywide Documents Access and Management System (ADAMS) in Accession No. ML18170A244.
In response to the DOE request, the NRC staff is providing preliminary comments in the areas that the NRC staff currently determined to be risk-significant and, therefore, likely to be the focus of a more detailed NRC staff review on the DOE document in the future. At this time, these NRC staff preliminary comments are provided to the DOE for information only. Thus, at this time, the NRC staff does not expect the DOE to provide any response to the NRC staff preliminary comments on the DOE document.
At the time that the following General Comments as well as Specific Observations and Comments were developed, the NRC staff had not yet reviewed the supporting documents to the DOE document. Therefore, the following does not contain all the NRC staff comments about the DOE document and its supporting documents.
NRC Staff General Comments (GCs):
GC-1: The NRC staff notes that, although Table 2 provides some measure of uncertainty, that table does not represent many of the sources of uncertainty associated with the long-term performance of engineered surface covers (e.g., uncertainty in long-term parameter values for a manmade drainage layer, uncertainty in the frequency of defects in the high-density polyethylene (HDPE) at thousands of years after site closure).
GC-2: The NRC staff notes that what is conservative from an infiltration aspect may not be conservative from an erosional aspect. Although the one-dimensional WinUNSAT-H model was considered conservative and two-dimensional simulations were considered unnecessary because of the shallow slope (<1.5 percent (%)) of the cover, a cover surface slope with a 3% to 5% gradient may produce more runoff and/or lateral drainage. That would potentially affect the stability of the cover.
GC-3: The NRC staff notes that NUREG/CR-0728 is a Contractor Report (CR).
Recommendations in a NUREG/CR are the recommendations of the contractor who wrote the NUREG/CR, rather than recommendations of the NRC.
GC-4: The NRC staff notes that Equation 2 (i.e., Giroud-Houlihan analytical solution) and the hydraulic conductivity value of the upper drainage layer are expected to be risk-significant. The NRC staff will review those areas in more detail.
07/2018 NRC Staff Preliminary Comments on DOE Document, SRRA107772-000009, Rev. A p. 2 GC-5: The NRC staff notes that reliance or decisions about model projections of infiltration at thousands of years after site closure should have multiple lines of reasoning to support the conclusions. That is especially true if most of the reasons are based on observations in much younger (i.e., decades old or less) engineered systems.
NRC Staff Specific Observations and Comments:
Regarding the relationship between percolation and precipitation:
SOC-1: Page 24 states: The relationship between percolation and precipitation also illustrates that the percolation rate levels off once the annual precipitation reaches approximately 1450 mm per year, apparently due to saturated hydraulic conductivity controlling the deep drainage through the profile. However; the NRC staff notes that if the results from modeling as shown in Figure 12 are correct, then the question remains as to where the additional precipitation water goes. The concern would then shift to erosion due to runoff and instability due to the lateral drainage.
Regarding the hydraulic conductivity of the geosynthetic clay liner (GCL):
SOC-2: The NRC staff notes that, although NUREG/CR-7028 is frequently referenced, the GCLs in Table 2 on page 26 were assigned a hydraulic conductivity value of 1x10-11 meters/second (m/s) based on a 14-year old Barnwell cover sample. That results in a percolation rate of less than one-fifth of that predicted for a GCL with a NUREG/CR-7028 value of 1x10-10 m/s. Additional justification may be required as to why the individual Barnwell sample is more representative than the range of covers evaluated in NUREG/CR-7028.
Regarding the assumption that fine-over-coarse interfaces at depth are stable over very long periods of time:
SOC-3: The NRC staff notes references to near-term examples in NUREG/CR-7028 as analogs for the stability of fine-over-coarse interfaces at depth over very long periods of time.
However, those examples are not ideal analogues due to the short time period of service of 5 years to 15 years.
SOC-4: The NRC staff notes the importance of the section in NUREG/CR-7028 documenting the exhumation of cover components after a relatively short service life. Fine particles were observed within the exhumed geotextiles, geonets, and geosynthetic drainage layers and apparently were transported there due to the migration of those fines.
SOC-5: Page 15 referenced fine-over-coarse layering at natural analog sites in the State of Washington. The NRC notes that those analog sites are not ideal analogs because they are natural and the layers observed were not artificially constructed. There is a potential that natural soil structures at a natural site may behave differently than soil in an engineered surface cover due to the disturbed characteristics of the soil within a cover.
SOC-6: Pages 15 to 17 described two long-term, man-made analog sites. The Tu-Dun tombs in China need further NRC staff review. The NRC staff notes that the Kyushu burial mound site in Japan used alternating layers of clay and loam, which is different than the current SDF cover design (i.e., drainage layer will be made out of sand).
07/2018 NRC Staff Preliminary Comments on DOE Document, SRRA107772-000009, Rev. A p. 3 Regarding the HDPE:
SOC-7: The NRC staff notes that Equation 3 uses thickness of the composite barrier (geomembrane/GCL), which was set at 10 millimeters (mm). However, the intended thickness for the HDPE geomembrane below the upper lateral drainage layer is 0.06 inches thick, the intended thickness for the GCL is 0.2 inches, and the combined intended thickness of the HDPE/GLC is 0.26 inches. That is equal to 6.6 mm, which is two-thirds as thick as the value used in Equation 3.
Regarding the GCL:
SOC-8: Page 19 states: The composite barrier, drainage layer, and overlying soils in the ECC
[engineered closure cap] promote accumulation of water above the composite barrier, and drying of soils below the composite barrier. However; the NRC staff notes that because the hydration of the GCL is an important component of its low hydraulic conductivity and the GCL is in contact with the closure caps Foundation Layer, it is important to know if the GCL would become drier and lose its effectiveness as a barrier as the Foundation Layer dries.
Regarding the Erosion Barrier:
SOC-9: Page 21 states: erosion barrier acts as a choke in the cover profile, forcing water redistribution to occur almost exclusively in the upper backfill layer and the topsoil. However; the NRC staff notes that would indicate that full saturation may occur above the erosion barrier and result in lateral flow, which would possibly affect the stability of the cover.
SOC-10: Page 20 states: No root systems were observed below the composite barrier at any of the covers evaluated in NUREG/CR-7028. However; the NRC staff notes that NUREG/CR-7028 did not examine covers with pine trees growing on them.
SOC-11: Page 21 states: erosion barrier remains saturated or nearly saturated at all times... However; the NRC staff notes that because roots accumulate in regions where water is more plentiful and readily extracted, Loblolly pine tap roots, which grow between 1 m and 2 m deep, could grow into the infill of the erosion barrier and expand in diameter as they get older.
After multiple generations of pine trees, the erosion barrier may lose its cohesion as an intact unit.