ML23017A084
| ML23017A084 | |
| Person / Time | |
|---|---|
| Site: | PROJ0734 |
| Issue date: | 04/18/2023 |
| From: | Hans Arlt NRC/NMSS/DDUWP/RTAB |
| To: | |
| Arlt H, Ridge A | |
| Shared Package | |
| ML23090A081 | List: |
| References | |
| eConcurrence 20230331-60017 | |
| Download: ML23017A084 (58) | |
Text
Technical Review: Hydrogeology, Groundwater Monitoring, and Far-Field Modeling of the U.S. Department of Energy 2020 Performance Assessment for the Saltstone Disposal Facility at the Savannah River Site Date April 18, 2023 Reviewer Hans Arlt, Sr. Risk Analyst, U.S. Nuclear Regulatory Commission
1.0 Purpose and Scope
The purpose of this U.S. Nuclear Regulatory Commission (NRC) staff Technical Review Report (TRR) is to document the NRC staff review of the hydrogeology, far-field monitoring, and groundwater monitoring used in the U.S. Department of Energy (DOE) 2020 Performance Assessment (PA) for the Saltstone Disposal Facility (SDF) at the Savannah River Site (SRS)
(DOE document SRR-CWDA201900001). The NRC staff performed this review to support a future decision about whether the DOE has demonstrated that radioactive waste disposal activities at the SDF are in compliance with the performance objectives of Title 10 of the Code of Federal Regulations (10 CFR) Part 61 at the SRS SDF pursuant to Section 3116(b) of the National Defense Authorization Act for Fiscal Year 2005 (NDAA). This technical review also supports NRC monitoring of the SDF under Monitoring Factor (MF) 7.01, Certain Risk-Significant Kd Values in Site Sand and Clay, MF 8.02 Groundwater Monitoring, MF 8.03, Identification and Monitoring of Groundwater Plumes in the Z-Area, MF 10.09, Kd Values for SRS Soil, MF 10.10, Far-Field Model Calibration, MF 10.11, Far-Field Model Source Loading Approach, and MF 10.12, Far-Field Model Dispersion. MFs 7.01, 8.02, 10.09, 10.10, 10.11, and 10.12 are detailed in the NRC staffs current 2013 plan for monitoring the SDF (available in the NRC Agencywide Documents Access and Management System (ADAMS) under Accession No. ML13100A113). MF 8.03 is detailed in a 2018 NRC letter to the DOE (ADAMS Accession No. ML18219B035) supplementing the current 2013 plan for monitoring the SDF.
The scope of this review included aspects of the hydrogeology that affect the projected dose to hypothetical members of the public at locations 100 m (328 ft) from the SDF boundary. In addition, the scope encompassed reviewing data and information obtained from groundwater monitoring at the SRS, and in particular, at the Z-Area. Also, the review scope included the NRC staffs evaluation of parameters and equations used in various models by the DOE to simulate flow and transport in the unsaturated zone and saturated zone. In this document, the NRC staff uses the term far-field to designate areas of review in both the unsaturated and the saturated zones as compared to the near-field which focuses on features, processes, and modeling activities below the closure cap and above the unsaturated zone.
2.0 Background
Groundwater monitoring at and near the DOE SRS SDF is described in the current 2013 NRC Monitoring Plan for the SDF) in MF 8.02, Groundwater Monitoring. Besides the 2020 SDF PA, in this TRR the NRC staff reviewed and evaluated information from various DOE documents, including annual SDF groundwater monitoring reports. This TRR relies on findings of the 2018 Groundwater Monitoring TRR (ML18117A494) for groundwater monitoring information that has been consistent and relatively unchanged since that TRR was issued.
Because the SDF and the SRS Tank Farms (i.e., F-Tank Farm (FTF) and H-Tank Farm (HTF))
are both located at the SRS, the General Separations Area (GSA) groundwater flow models have been, and are, used to construct local far-field models for each of those locations. In addition to the updated GSA models (i.e., from 2018 and 2021), the NRC staff reviewed previous versions of the GSA models (i.e., from 1999, 2004, and 2016). The NRC staff had previously documented its review of the 2016 and 2018 GSA models in a TRR (ML19277H550)
(referred to in this document as the 2016 and 2018 GSA Models TRR). For aspects of the GSA model that the DOE did not change from the versions reviewed in the 2016 and 2018 GSA Models TRR, and that are not specific to the SDF area, this document relies on findings from that TRR.
For the SDF far-field models, this review also was informed by the 2012 NRC Technical Evaluation Report (TER) (ML121170309), the NRC 2013 Monitoring Plan for the SDF, the 2015 DOE Responses to the NRC Request for Additional Information (RAI) on the DOE Fiscal Year 2013 Special Analysis Document for the SRS SDF (DOE document SRR-CWDA201400099),
the 2016 DOE Responses to the NRC RAI on the DOE Fiscal Year 2014 Special Analysis Document for the SRS SDF (DOE document SRR-CWDA201600004), and the 2019 NRC SDF Onsite Observation Visit (OOV) Report (ML19289A525).
3.0 Hydrogeology, Groundwater Monitoring, and Far-Field Modeling for the 2020 SDF PA 3.1 Groundwater Monitoring 3.1.1 Hydrogeology The hydrogeological system of the Z-Area and the SDF (see Figure 2 for a map of the Z-Area) consists of three major aquifers: (1) the Upper Three Runs Aquifer (UTRA); (2) the Gordon Aquifer; and (3) the Crouch Branch Aquifer (below the Crouch Branch Confining Unit; see Figure 1). The UTRA and Gordon Aquifer are expected to be impacted by radionuclides from the SDF (DOE document SRR-CWDA201400095; Figure B4). Contamination is not expected to affect the deeper Crouch Branch Aquifer because of an upward flow gradient between the Crouch Branch and Gordon Aquifers near Upper Three Runs Creek.
The Tan Clay Confinement Zone (TCCZ) has the potential to slow the downward groundwater flow and radionuclide transport within the UTRA. This local aquitard splits the UTRA into an upper part, referred to as the UTRA-Upper Aquifer Zone (UAZ), or UTRA-UAZ, and a Lower Aquifer Zone (LAZ), referred to as the UTRA-LAZ. The thickness and extent of the TCCZ vary and are important inputs into groundwater flow and transport models used by the DOE to help demonstrate expected compliance with applicable groundwater regulatory requirements. The DOE described that the methodology to identify the location of the TCCZ used pore pressure, resistivity, and corrected tip stress while the actual physical description of the sedimentary layer was much less frequently used (see 2016 NRC SDF OOV Report (ML16147A197)). The latest information on the thickness and extent of the TCCZ is incorporated in a model called the 2021 GSA model (DOE document SRR-CWDA202100065). The 2021 GSA model was created by the DOE through modifications to the 2018 GSA model and intended to achieve better alignment with the Z-Area hydrostratigraphy and contaminated plume observations downgradient of Saltstone Disposal Structure (SDS) 4. Section 3.3.1 of this TRR describes the 2021 GSA model in more detail.
Figure 1: Z-Area Cross Section (Figure 5 from SRNS-TR201400283). Location of Cross Section A-A found in Figure 3.
The Z-Area straddles a groundwater divide between Upper Three Runs Creek and McQueens Branch, causing lateral-moving water and any constituents in the water to discharge to either creek depending on the SDF source location. As the arrow in Figure 2 points out, the water table at SDS 1 is higher than at SDS 4, so that shallower groundwater flow is to the northeast.
Similarly, the water table is higher at SDS 6 and groundwater is assumed to flow to the north-northwest toward SDS 3 and SDS 5, while groundwater flow at SDS 2 is to the north-northeast.
The water table elevation at the Z-Area is approximately between 67 to 73 meter-above mean sea level (m-amsl) (220 to 240 ft-amsl). The DOE 2016 Z-Area SDF Groundwater Monitoring Report (DOE document SRNS-TR201600110) includes that the hydraulic conductivity (K) of the UTRA is 4.0 m/day (13 ft/day) with an effective porosity (n) value of 25 percent (%).
Calculations by the DOE provided a range of groundwater flow rates from 56.1 to 105 m/yr (184 to 344 ft/yr) for the years 2016 and 2015 in the Z-Area (DOE documents SRNS-TR201600110 and SRNS-TR201500300).
Figure 2: Approximate 2017 Water Table in Feet Above Mean Sea Level in the Z-Area (Figure 3 from SRNS-TR201700227) 3.1.2 Monitoring The DOE conducts an effluent monitoring and environmental surveillance program on an ongoing basis at the SRS. The data obtained through that program are summarized in an annual environmental report. A variety of environmental media, including groundwater; surface water; rainwater; air; vegetation; deer and hog meat; and soil, are monitored through that program. The most useful environmental data to monitor for NDAA monitoring at the SDF is the groundwater data from the Z-Area. The NRC staff expects that the groundwater will be the dominant pathway for evaluating long-term releases from the SDF and assessing compliance with 10 CFR Part 61 Subpart C Performance Objective (PO) §61.41 (Protection of the General Population from Releases of Radioactivity). In assessing compliance with PO 10 CFR Part 61 Subpart C PO §61.42 (Protection of Individuals from Inadvertent Intrusion), the NRC expects that groundwater pathways will dominate long-term doses to inadvertent human intruders.
Potential increases in the concentration of radionuclides or other saltstone indicators (e.g., nitrate, pH, alkalinity) in groundwater samples obtained near the SDF may indicate that radionuclides are leaching from the disposal structures and that the SDF may not be performing as expected. The usefulness of other environmental samples to SDF monitoring is somewhat limited because most of those samples are not obtained directly in the vicinity of the SDF and there are other potential sources of radioactivity at SRS, which makes it difficult to determine
whether any observed concentration increases are attributable to the waste disposed at the SDF.
The groundwater monitoring network has expanded and will expand with the construction of additional disposal structures. The Saltstone Production Facility is permitted at SRS as a wastewater treatment facility per the South Carolina Department of Health and Environmental Control (SCDHEC) regulations, and the SDF is permitted as a Class 3 Landfill per SCDHEC regulations (SCDHEC, 2008). In accordance with the Z-Area SDF Class 3 Landfill Permit, the monitoring wells are sampled for specific constituents and parameters (see Table 1 in the 2018 Groundwater Monitoring TRR) on a semiannual and biennial basis. If a 30 pCi/L threshold is exceeded by a well sample, then the DOE would resample the same well and an applicable background well (Well ZBG 1 or Well ZBG 15D) within 30 days for the specific constituents.
Background concentrations are based on historical data from Wells ZBG 1 and ZBG 15D, which are upgradient of the SDF but downgradient of several other SRS facilities. If any contingent constituent is above maximum background well concentrations, then the SCDHEC will add it to the list of constituents in the Class 3 Landfill Permit.
Groundwater monitoring has been conducted at the Z-Area SDF since 1987 (DOE document WSRC-TR200500257). The Z-Area SDF started operations and disposing of low-level waste by pumping saltstone to SDS 1 in June 1990 (DOE document WSRC-TR950227). The DOE began to monitor the groundwater with two wells located in the UTRA-UAZ (designated Wells ZBG 1 and ZBG 2) to determine baseline chemistry. Wells ZBG 3, ZBG 4, and ZBG 5 were constructed in 2003 with open screens in the UTRA-LAZ downgradient of SDS 4. In addition, three wells (i.e., ZBG 6, ZBG 7, ZBG 8) were installed immediately downgradient of SDS 1 in the UTRA-UAZ in 2007. Since then, additional monitoring wells have been constructed (see Figure 3), most of them in the UTRA-LAZ. Well ZBG 2C was installed in the UTRA-LAZ adjacent to shallow Well ZBG 2, when contaminated water was observed at Well ZBG 2.
Ultimately, Well ZBG 2 was abandoned due to DOE concerns that it may act as a pathway for contamination from the UTRA-UAZ into the UTRA-LAZ. Well ZBG 2 was replaced by the shallow Well ZBG 2D, which does not dissect through the TCCZ. The current groundwater monitoring system does not include groundwater monitoring wells in the Gordon Aquifer.
A key aspect of the DOE groundwater monitoring program is the placement of the wells (see Figure 3). For the monitoring wells to provide useful information, they must be located downgradient of the disposal structures and must be close enough to the disposal structures to see radionuclides or other indicators that may leach from the saltstone waste form. Similarly, it is important for the wells used to obtain information regarding the natural groundwater composition (i.e., background wells) to be upgradient of the disposal structures. The groundwater divide on the SDF may complicate the assessment of the well locations, especially because there is some uncertainty in the location of the divide and the location can change with changes to infiltration.
Figure 3: Groundwater Monitoring Wells in the Z-Area and Z-Area Cross Section A-A (Figure 2 from SRNS-TR202100966) 3.1.3 Summary of Current Groundwater Monitoring Results The NRC staff focuses on several areas of the groundwater monitoring program. Specifically, the NRC staff reviews: (1) if the current wells are adequate to assess whether leaching from saltstone has occurred; (2) if the locations of the background wells are adequate to assess leaching from saltstone by comparing background concentration values with concentration values from wells downgradient of disposal structures; (3) if the number and the locations of the groundwater monitoring wells are adequate; (4) if screen well openings are at appropriate elevations to capture contaminants in the aquifer; and (5) if groundwater monitoring data show any increase in the concentration of radionuclides or saltstone indicators. The NRC staff considers historical concentrations measured in the groundwater and measured concentrations in upgradient wells when evaluating the appropriateness of the background levels.
The SRS Z-Area was not used as an SRS processing site, so that, until the first constructed disposal structure was filled with saltstone, no contaminants, besides background contaminants, were present in the Z-Area groundwater. After SDS 1, and in particular, after SDS 4 had saltstone poured into them, contaminants from the contents of those disposal structures have entered the groundwater. Numerous groundwater samples with these contaminants were collected by various means. Higher nonvolatile gross beta concentration values were found in
four of the twelve direct push technology (DPT) samples from the 2015 Z-Area groundwater characterization study (DOE document SRNS-RP201500902) and samples from monitoring Wells ZBG 2, ZBG 2D, and ZBG 20D included nonvolatile gross beta concentrations greater than 30 pCi/L over multiple years after 2012. ZBG 2D and five of the twelve DPT samples exceeded the groundwater protection standard of 15 pCi/L for gross alpha in 2015 while ZBG 3 yielded a radium (Ra)-226 result of 41.1 pCi/L in 2005, although the Ra-226 result may have been a sampling or laboratory error (DOE document SRNS-RP201500902).
The DOE reported (DOE document ESH-WPG200600132) that SDS 1 experienced cracks from construction and operational events dating back to 1988. Many of the cracks were sealed on the outside, and soil in areas where leakage occurred has been isolated and partially remediated; however, elevated hydrogen (H)-3 values were observed in Well ZBG 6 samples in 2007 and 2008. These samplings did not reveal additional evidence of new releases from the disposal structures. H3, or tritium, in Well ZBG 6 may be evidence that contamination from a 1994 leak at SDS 1 has reached the water table (DOE document WSRC-TR200800001). The DOE did not provide a delineation of a tritium plume in the UTRA-UAZ.
The DOE indicated that bleed water and contaminated rainwater weeping from Cell G of SDS 4 in the 1990s (DOE document SRR-CWDA201600052; page 84) had reached the unsaturated zone and saturated zone (DOE document SRR-CWDA201600004). The DOE also indicated that the release was responsible for the elevated technetium (Tc)-99 and nitrate concentrations measured in Well ZBG 2 starting in 2012. As described in the 2015 NRC OOV Report (ML15236A299), in 2011, precipitation accumulating on top of SDS 4 may have flowed through cracks in SDS 4, and again contaminated the ground surface. The DOE remediated the newly contaminated soil around SDS 4 and indicated that the remaining contaminants were immobile and that the contaminants from the 2011 leak did not reach the saturated zone and were not present under SDS 4. A stormwater runoff system was built after the 2011 precipitation event, which included a drain system just under the surface that encircles SDS 4 and channels the water directly to the sedimentation basin.
In 2014, higher concentration data from Well ZBG 2 initiated the development of a characterization plan for nonvolatile gross beta, Tc-99, and nitrates in the groundwater in Z-Area (DOE document SRNS-RP201401214). The DOE indicated that it is important to delineate the top of the TCCZ to identify features (depressions and/or troughs), where the contaminated water is most likely to exist prior to collecting groundwater samples (DOE document SRNS-RP201500902). The DOE implemented a groundwater characterization plan in July 2015 that included cone penetrometer test (CPT) pushes and DPT groundwater samples to characterize the extent of the groundwater plume. Based on data from the SRNS-RP201500902 characterization data report, Tc-99 and nonvolatile gross beta groundwater plumes were identified as shown in Figures 4 and 5. The only contaminants identified by the DPT groundwater characterization samples were nonvolatile gross beta, Tc-99, and nitrates.
Figure 4: Approximate Extent of the Tc-99 Groundwater Plume and Contours of the Top of TCCZ in Z-Area (Figure 22 from SRNS-RP201500902)
Figure 5: Approximate Extent of the Nonvolatile Gross Beta Groundwater Plume and Contours of the Top of TCCZ in Z-Area (Figure 21 from SRNS-RP201500902)
At SRS, nitrogen in the groundwater is primarily in the form of nitrate because the groundwater is typically well oxygenated, especially in the UTRA-UAZ. The 2021 maximum nitrate groundwater concentration of 8.77 mg/L at ZBG 2D exceeds the maximum concentrations in background Wells ZBG 1 and ZBG 15D. Ra-226, with a half-life of about 1,600 years, and its daughter radionuclides of lead (Pb)-214 and bismuth (Bi)-214, contribute to the gross alpha activity in groundwater. In 2021, the maximum gross alpha groundwater concentration of 3.14 pCi/L at ZBG 6 was greater than the historic maximum of 1.48 pCi/L for background Well ZBG 15D.
Concentrations of contaminants have overall decreased since 2015 at Well ZBG 2D, where the highest concentrations of the SDS plume have historically been measured, in addition to the now decommissioned Well ZBG 2. However, the general downward concentration trend has not been consistent, but instead values increase and decrease every two to four years as can be seen in Figure 6. In 2021, as compared to 2020 results, groundwater at Well ZBG 2D again showed increases in nonvolatile beta activity (44.6 pCi/L), nitrate concentration (8.77 mg/L) and specific conductance (90 S/cm). Results for Tc-99 activity also had higher concentrations for 2021: 117 pCi/L up from 51.3 pCi/L in 2020.
Figure 6: ZBG 2 / ZBG002D Conductivity, Nonvolatile Beta, Tc-99, and Nitrate Groundwater Trends (Figure 9 from SRNS-TR202100966)
Surface water and sediment contamination was detected in 2011 in Sedimentation Basin No. 4 (DOE document SRNS-TR201700387). Subsequent DOE probe surveys identified radiation above background levels in the outfall culvert area. Concerns of potential groundwater contamination from Sedimentation Basin No. 4 initiated the installation of a well in the UTRA-
UAZ (ZBG 16D) to monitor potential saturated conditions in the UTRA-UAZ and a deeper well in the UTRA-LAZ (ZBG 16C) to monitor the UTRA-LAZ. The screen zone for Well ZBG 16D is positioned on top of the TCCZ where saturated conditions have not been encountered since the installation of this well.
3.2 Flow and Transport in the Unsaturated Zone Below the Disposal Structures The following sections focus on the unsaturated soils of the vadose zone models. The lower vadose zone soils are the deepest component of the Vadose Zone Model and found beneath each disposal structure. The lower vadose zone soils, or native soils as they are labeled in the 2020 SDF PA, form one layer material of the 27 distinct material types (or mtyp zones) defined in PORFLOW to represent different materials and to facilitate modeling of different flow paths.
Other Vadose Zone Flow Model layer material types include the lower backfill on top of the model and the various disposal structure components. In this TRR, the lower vadose zone soil between the bottom of the disposal structure mud mat and above the water table is defined as the unsaturated zone.
3.2.1 Vadose Zone Models Used in the 2020 SDF PA The deterministic Vadose Zone Flow Model used the infiltration rates from the Closure Cap Model and the hydraulic degradation rates from the Cementitious Degradation Model to develop flow rates through disposal structure concrete, saltstone, backfill, and native soils above the water table to estimate unsaturated flow rates using PORFLOW software. The leakage rates through the HDPE/GCL composite barrier layer in the closure cap served as the basis for the infiltration rate input for the Vadose Zone Flow Model. Short-term parameter variations, such as seasonal precipitation, were not simulated. Two-dimensional flow fields were generated separately for SDS 1, SDS 4, SDS 2, SDS 6, SDS 7, and SDS 9, which can represent every planned disposal structure (e.g., model results for SDS 7 represent SDS 8, SDS 10, SDS 11, and SDS 12). The infiltrating water flows through, and exits, the disposal structures, and then, enters the unsaturated soils above the water table as percolation. The modeled thicknesses of these unsaturated soils between disposal structures and water table varies between 15 m (48 ft) for SDS 1 and 7 m (23 ft) for SDS 9. The unsaturated zone consists of lower vadose zone soils representing undisturbed sediments between an elevation of 80.5 m (264 ft) and the water table below. The lower boundary of the Vadose Zone Flow Model coincides with the water table where the pressure head is set to zero. Modeled flow fields were generated by running transient flow simulations for specific time intervals until steadystate conditions were approached. The outputs were combined, and the resulting flow fields were then forwarded as input to the Vadose Zone Transport Model.
As with the Vadose Zone Flow Model, the deterministic Vadose Zone Transport Model was constructed using PORFLOW software. The Vadose Zone Transport Model simulates radionuclide releases from the saltstone. It also simulates radionuclide transport through the saltstone, concrete components of the disposal structures, and the unsaturated zone of the native soils. The rates of release and transport are based on the flow fields generated by the Vadose Zone Flow Model for the specific disposal structures and on the transport properties for the individual contaminants. Processes such as sorption, precipitation/dissolution reactions, diffusion, and advection control the release and transport of saltstone contaminants. Using these processes, the Vadose Zone Transport Model provides estimated fluxes (i.e., the timevariant rate of mass exchange) of contaminants at the water table, which is then used as input to the Aquifer Transport Model. The bottom boundary condition (the outflow) of the Vadose
Zone Transport Model at the unsaturated zone/water table boundary is a zero-concentration boundary condition.
3.2.2 Hydrostratigraphy The 2020 SDF PA divided the vadose zone at the SDF into two regions, the upper vadose zone and the lower vadose zone, using an elevation of 80.5 m (264 ft) above mean sea level as a boundary between the two. The upper vadose zone consists of finergrained sediments while the lower vadose zone has a higher sand content. As previously mentioned, the depth between the base of the disposal structure and the elevation of the water table (obtained from the 2018 GSA groundwater model (DOE document SRNLSTI201800643)) can vary; however, the modeled depth of the vadose zone was assumed to be the same for SDSs 7, 8, 10, 11, and 12 (i.e., 11 m [36 ft]) and different for SDS 9 (i.e., 7.0 m [23 ft])
In the DOE Response to the NRC Second Set of RAI Questions for the Review of the SRS SDF PA under Clarifying Comment (CC)-10 (DOE document SRR-CWDA202100072), the DOE described the prevalence of upper vadose zone and lower vadose zone soils in the planned SDF. During the construction of the disposal structures, the walls are completely exposed, with no surrounding backfill or upper vadose zone soils, but are then surrounded by backfill material during the construction of the engineered surface cover. This compacted backfill is intended to consist basically of the upper vadose zone soils. While the backfill material properties were developed based on the material properties of the upper vadose zone soils, the upper vadose zone soils are not explicitly used in any of the SDF PA models. This is consistent with the vadose zone modeling which assumes all disposal structures are above the native soil that have been assigned the properties of the lower vadose zone and surrounded by compacted backfill.
To evaluate the potential impacts from climate change, a wetter climate scenario and a drier climate scenario were developed and evaluated with the SDF GoldSim Model.
To support the climate change scenarios, unsaturated zone thicknesses were modified for each of the disposal structures (i.e., the modeled water table was adjusted lower during a dry climate and upward during a wetter climate). These thicknesses were adjusted using data from the now decommissioned monitoring Well ZBG 2, which provided the longest range of monitoring within the SDF boundaries. Based on the analysis presented in the DOE document SRRCWDA2019 00027, the unsaturated zone was thickened by 2.3 m (7.6 ft) for the dry climate scenario and decreased by 2.4 m (8 ft) for the wet climate scenario while the saturated zone thickness was adjusted to reflect these changes in the unsaturated zone.
3.2.3 Unsaturated Zone Material Properties Because the properties of the upper vadose zone are more clayey and would tend to retard contaminant transport compared to backfill or lower vadose zone materials, the DOE assumed soil with properties similar to compacted backfill will surround the disposal structures and that all of the soil underneath the disposal unit is similar to lower vadose zone soil (rather than upper vadose zone) material. The hydraulic properties of the unsaturated and saturated soils used in the 2020 SDF PA are listed in Table 1. Vadose zone material properties are assumed to be constant, and do not change with time.
Table 1: Hydraulic Properties of the Soils Surrounding and Below the Disposal Structures Used in the 2020 SDF PA Models (Table 4.32 from SRR-CWDA201900001)
Parameter Sand Drain Backfill Upper Vadose Lower Vadose Gravel Saturated Effective Diffusion Coefficient De (cm2/sec) 8.0E06 5.3E06 5.3E06 5.3E06 9.4E06 Average Total Porosity (%)
41.7 35 39 39 39 Average Dry Bulk Density (g/cm3) 1.55 1.71 1.65 1.62 1.82 Average Particle Density (g/cm3) 2.66 2.63 2.70 2.66 2.60 Saturated Horizontal Hydraulic Conductivity (cm/s) 5.0E02 7.6E05 6.2E05 3.3E04 1.5E01 Saturated Vertical Hydraulic Conductivity (cm/s) 5.0E02 4.1E05 8.7E06 9.1E05 1.5E01 Most unsaturated zone material properties values were originally obtained from the DOE document WSRCSTI200600198, Hydraulic Property Data Package for the EArea and ZArea Soils, Cementitious Materials, and Waste Zones including moisture characteristic curves. When compared to saturated soils, water will be attracted more to soil that is unsaturated since space is available for additional water. When this suction head or water potential is contrasted with the existing water content, moisture characteristic curves, or MCCs, can be developed. The MCCs for the soil materials used in the vadose zone models were described in the DOE document SRNLSTI201800652 and developed using the van Genuchten function and parameters. They were presented in the 2020 SDF PA and as shown in Figure 7.
Figure 7: Moisture Characteristic Curves for the Soils Surrounding and Below the Disposal Structures Used in the 2020 SDF PA Models (Figure 4.31 from SRR-CWDA201900001)
3.2.4 Percolation The DOE stated in the 2020 SDF PA that there is not much difference in the volumetric flow rates through the different modeled material zones indicating that flow is predominately downward. An example figure from the 2020 SDF PA, see Figure 8, shows that the volumetric flow rates through the various materials within the model closely resemble the infiltration rates entering the Vadose Zone Flow Model (uppermost material zone is the backfill). Together with the increasing hydraulic conductivity (at 2000 years) of the upper composite barrier layer, volumetric flow rates through the disposal structure and saltstone grout increase in time after the upper lateral drainage layer decreases in hydraulic conductivity (at 500 years) since water is no longer able to laterally exit the cover system as quickly via the drainage layer. The upper composite barrier layer increases in hydraulic conductivity (at 2000 years). Figure 8 also indicates that flow rates appear greater for the backfill and native soil which may be due to the fact that the volumetric flow rates alongside the disposal structure are included in the summation while the other component flow rates are restricted to flow within the disposal structure.
Figure 8: Volumetric Flow Rates for the Compliance Case Through SDS 7 Materials (Figure 4.472 from SRR-CWDA201900001)
3.2.5 Sorption Coefficients As discussed in Section 3.2.1 of this TRR, processes such as sorption (modeled with partition coefficients), precipitation/dissolution reactions (modeled with solubility limits), diffusion, and advection control the release and transport of saltstone contaminants. In both the 2009 SDF PA and the 2020 SDF PA, the DOE modeled radionuclide sorption to soils with linear sorption coefficients (Kd values). Because the NRC staff reviewed those soil sorption coefficients in the 2012 NRC TER, this review focuses on the following: (1) sorption values for the radionuclides that make a significant contribution to the projected dose to an offsite member of the public in the 2020 SDF PA (i.e., iodine (I) and technetium (Tc)), (2) differences between the 2009 and 2020 SDF PA modeling of radionuclide sorption to soils, and (3) issues highlighted in the 2012 NRC TER. Although isotopes of plutonium, tin, and americium contributed to dose in certain intrusion scenarios, the exposure pathways that led to dose from those radionuclides in those intrusion scenarios did not rely on radionuclide transport in groundwater; therefore, the soil sorption coefficients for those radionuclides did not significantly affect the projected dose.
In the 2012 NRC TER, the NRC staff concluded that the DOE process for selecting radionuclide sorption coefficients was acceptable. The DOE used site-specific data when possible and relied on literature data when site-specific data were not available. When site-specific data were not available for a risk-significant radionuclide, the DOE made conservative assumptions. The DOE used the same process for selecting soil sorption coefficients in the 2020 SDF PA.
For iodine, the DOE based the sorption coefficient on measurements from the SRS F-Area (DOE document SRNL-STI201200518). To show the applicability to Z-Area, the DOE provided a comparison of soil characteristics from F-Area and Z-Area (DOE document SRR-CWDA202100072). That comparison showed that, in general, the soil in Z-Area contains more clay and less sand than the soil in F-Area. Based on that observation, the DOE determined that the sorption coefficients for iodine based on measurements in F-Area soils were likely to underestimate the sorption of iodine on Z-Area soil because, like most elements, iodine generally sorbs more to clay than it does to sand.
In addition to soil characteristics, the DOE also considered the expected chemical form of iodine in the site soil. In the DOE document SRNL-STI202100518, SRS subsurface clayey and sandy soils had a mean measured sorption coefficient of 4.6 mL/g when spiked with iodide and 20 mL/g when spiked with iodate (IO3-). In the DOE document SRNL-STI202100518, the DOE indicated that subsurface conversion of iodide to iodate or organo-iodine could explain why some measured sorption coefficients in SRS soils are greater than literature values of iodide sorption to soil.
For technetium, the sorption coefficients were all either within 20% of the values in the 2009 PA or were less than the values the DOE used in the 2009 PA. The NRC staff previously reviewed the values used in the 2009 PA and found them to be sufficiently supported. Values less than those values would tend to increase radionuclide transport to an offsite member of the public.
In addition to iodine and technetium, in the 2012 NRC TER the NRC staff also reviewed soil sorption coefficients for strontium (Sr), radium (Ra), and selenium (Se) because the NRC staff had determined that those radionuclides could have a significant effect on the projected dose in the 2009 SDF PA. Because of changes in the modeled inventory of Ra, which the NRC staff reviewed in the NRC TRR on inventory for the 2020 SDF PA (ML23017A087), the NRC staff does not expect Ra to make a significant contribution to dose to a human receptor. In the 2012
NRC TER, the NRC staff determined that the soil sorption coefficients for Tc, I, Sr, and Ra had an adequate technical basis. In contrast, the NRC staff concluded the soil sorption coefficients for selenium (Se) did not have an adequate technical basis. The DOE did not change the soil sorption coefficients used for Se in the 2020 SDF PA.
In the 2012 NRC TER, the NRC staff described that the 2009 SDF PA models did not account for the effect that the alkalinity of grout and disposal structure concrete leachate could have on radionuclide sorption to vadose zone soil. To address the potential chemical effects of the leachate on soil sorption coefficients in the 2020 SDF PA, the DOE implemented leachate impact factors for soil in the vadose zone (i.e., clayey backfill soil and sandy vadose zone soil).
The DOE did not use leachate impact factors to model radionuclide sorption in the saturated zone because the DOE expects that groundwater flow in the aquifer will dilute the alkaline leachate enough that the leachate will not affect radionuclide sorption to aquifer soil. The leachate impact factors are the ratio of sorption coefficients measured in leachate-impacted and non-impacted soil:
,=
where represents each element.
In the 2020 SDF PA, the DOE based leachate impact factors on a study performed with Hanford sediments (DOE document PNNL16663). Although the DOE subsequently conducted a study with SRS soil (DOE document SREL Doc. No. R130004), that study did not recommend any changes to the leachate impact factors from the Hanford study because limitations of the analytical methods the DOE used in the SRS study limited the conclusions the authors could draw from the data.
In the Compliance Case for the 2020 SDF PA, the DOE modeled all sorption coefficients in the vadose zone as being leachate-impacted for the entire model duration. Section 4.3.2 in the 2020 SDF PA indicated that the DOE modeled leachate impacts lasting until the saltstone transitions from a chemically reduced to a chemically oxidized state (i.e., an Eh transition). In contrast, Section 4.4.3.3.3 in the 2020 SDF PA indicated that the DOE modeled leachate impacts lasting until the saltstone and disposal structure concrete transition from Region III to Region IV conditions (i.e., a pH transition). Although Table 4.462 in the 2020 SDF PA shows that the DOE expects the two transitions to occur at different times, the difference is not risk-significant because both transitions occur in saltstone grout more than 50 million years after closure in the Compliance Case.
Table 2 lists Kd values used in transport modeling for the two basic types of soil in the Z-Area (sandy and clayey) and for soils effected by leachate originating from saltstone.
Table 2: Modeled Kd Values for Backfill, Clayey Soils, and Sandy Soils in the 2020 SDF PA (Table 4.34 from the 2020 SDF PA)
Element or Ion Backfill or Clayey Soils (mL/g)
Leachate-Impacted Clayey Soils (mL/g)
Vadose Zone or Sandy Soils (mL/g)
Leachate-Impacted Sandy Soils (mL/g)
Ac 9,000 10,000 1,000 2,000 Ag 30 100 10 30 Al 1,000 2,000 1,000 2,000 Am 9,000 10,000 1,000 2,000 As 200 300 100 100 B
- 0 0
0 0
Ba 100 300 20 50 C
400 2,000 10 50 Cd 30 90 20 50 Cf 9,000 10,000 1,000 2,000 Cl 8
0.8 1
0.1 Cm 9,000 10,000 1,000 2,000 Co 100 300 40 100 Cr 1,000 1,000 400 600 Cs 50 50 10 10 Cu 70 200 50 200 Eu 9,000 10,000 1,000 2,000 F
8 0.8 1
0.1 Fe 400 600 200 300 H
0 0
0 0
Hg 1,000 3,000 800 3,000 I
3 0.3 1
0.1 K
30 30 5
5 Mn 200 300 20 20 Mo 1,000 1,000 1,000 1,000 N
8 0.8 1
0.1 NO2
- 0 0
0 0
NO3
- 0 0
0 0
Nb 1,000 1,000 1,000 1,000 Ni 30 100 7
20 Np 9
10 3
5 Pa 9
10 3
5 Pb 5,000 20,000 2,000 6,000 PO4
- 0 0
0 0
Pt 30 100 7
20 Pu 6,000 10,000 650 1,000 Ra 200 500 30 80 Rn 0
0 0
0 Sb 3,000 4,000 3,000 4,000 Se 1,000 1,000 1,000 1,000 Sm 9,000 10,000 1,000 2,000 Sn 5,000 20,000 2,000 6,000 SO4
- 0 0
0 0
Sr 20 50 5
20 Tc 1.8 0.2 0.6 0.06 Th 2,000 4,000 900 2,000 U
400 1,000 300 900 Zn 30 90 20 50 Zr 2,000 4,000 900 2,000
Table 3 provides a comparison between the soil sorption coefficients the DOE used in the 2009 SDF PA and the 2020 SDF PA. Table 3 uses the same values from the 2009 PA to represent leachate-impacted and non-impacted sorption coefficients because the 2009 SDF PA did not use leachate impact factors.
Table 3. Comparison Sorption Coefficients in SRS Soils Used in the 2009 SDF PA and 2020 SDF PA Backfill (clayey soil)
Leachate-Impacted Backfill Vadose Zone (sandy soil)
Leachate-Impacted Vadose Zone Soil Decreased by more than 20%
Ag, Al, Cs, K, Np, Pa, Zn Ag, Cs, I, K, Np, Pa, Tc, Zn Ag, Al, Cs, K, Zn Ag, Cs, K, Tc, Zn Within 20% of 2009 value Ac, Am, B, Cf, Cm, Cu, Eu, Fe, Hg, Mn, Ni, Pb, Pu, Rn, Sb, Se, Sm, Sn, Sr, Tc, Th, Zr, Ac, Am, B, Cf, Cm, Eu, Rn, Se, Sm Ac, Am, B, Cf, Cm, Cu, Eu, Fe, Hg, Ni, Pb, Rn, Sb, Se, Sm, Sn, Sr, Tc, Th, Zr B, Rn, Se Increased between 20%
and 100%
U Al, Fe, Mn, Pu, Sb, Th, Zr Mn, U Ac, Al, Am, Cf, Cm, Eu, Fe, Mn, Sb, Sm Increased between a factor of 2 and a factor of 10 Ba, Cd, Co, I Cd, Co, Cu, Hg, Ni, Pb, Sn, Sr, U Ba, Cd, Co, No, Pa, Pu, Ra Ba, Cu, Hg, Ni, Np, Pa, Pb, Pu, Sn, Sr, Th, U Increased between a factor of 10 and a factor of 100 Cr Cr Cr Cr Increased from a value of 0 mL/g in the 2009 SDF PA (2020 SDF PA value in mL/g)
C (400), Cl (8), F (8), Mo (1,000), Nb (1,000), Pt (30)
C (2,000), Cl (0.8), F (0.8),
Mo (1,000),
Nb (1,000), Pt (100)
C (10), Cl (1),
F (1), Mo (1,000), Nb (1,000), Pt (7),
Pt (20)
C (50), Cl (0.1), F (0.1),
Mo (1,000),
Nb (1,000), Pt (20)
The DOE conducted two deterministic sensitivity analyses with the SDF GoldSim Model to provide information about the risk significance of soil sorption coefficients for the SDF. In one analysis, the DOE reduced all sorption coefficients in the model (i.e., including the disposal structure concrete, saltstone, and all soil types) by a factor of 100. That analysis showed that the reduction had a small effect on the projected dose to an off-site member of the public (Figure 9 in this TRR). In another analysis, the DOE kept the sorption coefficients in the saltstone grout and disposal structure concrete unchanged and reduced all soil sorption coefficients by a factor of 1 x 1030. That analysis demonstrated that essentially removing the effect of radionuclide sorption to soil had a small effect on the projected dose to an offsite member of the public (Figure 10 in this TRR).
Figure 9: Deterministic results of reducing all sorption coefficients (Kd values) in the SDF GoldSim Model (Figure 5.835 in the 2020 SDF PA)
Figure 10: Deterministic results of reducing soil sorption coefficients (Kd values) in the SDF GoldSim Model (Figure 5.837 in the 2020 SDF PA)
The DOE found that most of the differences in Figures 9 and 10 could be attributed to the change in the iodine sorption coefficients in soil. The DOE showed the risk significance of assumptions about the iodine sorption coefficients in subsurface soil by reducing the modeled sorption coefficients for iodine in sandy and clayey soil, with and without leach impact factors, to approximately one-third of their Compliance Case values. That sensitivity analysis showed that
significantly reducing the sorption coefficients for iodine in leachate-impacted and non-impacted sandy and clayey soil in the Compliance Case increased the projected peak dose to a member of the public within 10,000 years of site closure by 10%.
To demonstrate the effect of applying the leachate impact factors, the DOE compared the results of applying the leachate impact factors to no soil or all soils to the Compliance Case. As shown in Figure 11, applying the leachate impact factors to all site soils had a negligible effect on the projected dose. In contrast, removing the leachate impact factors from all soils decreased the projected dose by 40%.
Figure 11: Effect of leachate impact factors (Figure 5.839 in the 2020 SDF PA)
3.2.6 Uncertainty
Probabilistic Analyses In the 2020 SDF PA, the DOE performed probabilistic sensitivity analyses and provided both partially ranked correlation coefficients (PRCCs) and standardized rank regression coefficients (SRRCs). Probabilistic sensitivity analyses help identify parameter distributions that have the greatest influence on the PA results. PRCCs show the degree of this influence, specifically between an independent variable and a specific model result, after the independent variables are analyzed one at a time by isolating and ignoring the combined effects of the other independent variables. When ranktransformed data is analyzed, the PRCCs are calculated with their ranks rather than with the original values for the variables after the original values were replaced by their ranks. The degree of a parameters influence on the results is frequently a function of time, and PRCCs can help bring the importance of a parameter, and the parameters importance over time, in focus.
Stepwise rank regressions help determine the effects of uncertain inputs on analysis results and involves ranktransformed variables rather than the original values. SRRCs are calculated at fixed time (e.g., the time of the peak of the means) unlike PRCCs which are computed as a function of time. All SDF model output variables from the 2020 SDF PA generally used the same times for analysis to facilitate comparisons between the sensitivity analysis results across output variables.
The PRCCs for total dose to the member of the public (MOP) at any sector are shown in Figure 12. Initially, that is starting at 100 years and lasting for over 1,000 years, the thickness of the unsaturated zone (Vadosethick_SDU9) is the most dominant parameter with a strong negative correlation (PRCC = 0.97) to the total dose values. These results are based in the relatively thin unsaturated zone thickness of 7.0 m (23 ft) beneath SDS 9, plus or minus 5.0 m (16.4 ft), as discussed in Section 3.2. Simulated releases from SDS 9 reach the saturated zone and the 100meter boundary relatively quick due to the short distance the radionuclides need to travel in the unsaturated zone. However, the influence of the unsaturated zone thickness becomes negligible as the releases from the other disposal structures begin to reach the 100 meter boundary after 5,000 years.
The partition coefficient for iodine in leachateimpacted sandy soils (labeled Kd_SandL_I in Figure 12) is also important within the first 1,000 years. However, it is not as dominant as Vadosethick_SDU9 and quickly decreases in importance after 1,000 years.
Figure 12: The PRCCs of the Total MOP Dose for Any Sector Demonstrating the Importance of the Thickness of the Unsaturated Zone (Figure 5.710 from SRR-CWDA201900001)
The stepwise regression analysis confirms the relative importance of the unsaturated zone thickness as seen in Table 5.78 in the 2020 SDF PA. The first four or five variables effectively dominate any influence over the dependent variable at the time analyzed and Vadosethick_SDU9 has the highest SRRC for the dose results analyzed at 1,000 years.
3.2.7 Uncertainty
Deterministic Sensitivity Analyses Deterministic sensitivity analyses also examined the sensitivity of unsaturated zone thicknesses to the results. As discussed in Section 3.2 of this TRR, to support the climate change scenarios, unsaturated zone thicknesses were modified for each of the disposal structures. The approach assumes that changing climate conditions impacts more than just the unsaturated/saturated thicknesses, the other parameters associated with the parametric flow cases (cementitious degradation rate, saturated hydraulic conductivity, and initial saltstone hydraulic conductivity) were also modified to capture more extreme flow conditions in addition to different infiltration rates and to develop flowrate multipliers for the saturated zone flow rates. The results of the analysis show that over the compliance period, the peak dose for the wetter climate case is approximately 3.7 times higher than the peak dose for the Compliance Case, and the peak dose for the Compliance Case is approximately 3.5 times higher than the peak dose for the drier climate case (Section 5.8.2.6 in the 2020 SDF PA).
The assumed homogeneity was also tested in a sensitivity case involving an impermeable soil layer within the unsaturated zone on which a perched zone develops (feature, event, and process (FEP) 5.2.09). Results from this case show that slightly lower releases occur due to the added layer functioning as a diffusion barrier when compared to the original lower vadose zone sediments (Section 5.8.9.4 in the 2020 SDF PA).
Another sensitivity case evaluated general Kd sensitivities including the Kd values of the native soils by decreasing all the Kds by a factor of 100. Multiplying all Kds by 0.01 resulted in increasing the peak from 1.8 mrem/yr to 2.1 mrem/yr and shifting the occurrence of the peak from 10,000 years to 9,920 years. The minor influence this change in the Kd values has to do with the dominance of I129 and Tc99 on dose, which have small initial Kd values and are already easily transported to the receptor. Because of this dominance, reducing other less-significant radionuclide Kd values has minimal impact on SDF performance.
3.3 Flow and Transport in the Saturated Zone Flow through the saturated layers of the SRS is modeled in PORFLOW via the 2018 GSA model as documented in the DOE document SRNLSTI201800643. The 2004 GSA model was updated in 2016 and then again in 2018 due to additional field data acquired since the mid-1990s and to the maturation of optimization software available for model calibration (DOE document SRNL-STI201700008). The model was updated once more in 2021 whereby the tops of the TCCZ and the UTRA-LAZ were modified near and downgradient of SDS 4 using the current hydrostratigraphic picks. Consequently, the 2021 GSA model incorporates the most current data on elevation and thickness of the TCCZ. Moisture characteristic curves were also modified to increase the downward component of vadose zone flow in 2021 GSA model simulations (DOE document SRR-CWDA202100065).
Since the GSA model is regional in nature (i.e., simulates flow in the entire GSA), numerous SRS PAs incorporate those parts of the GSA model that cover the facility and area being assessed (Figure 13 in this TRR). Precipitation, soil properties, hydrologic stratigraphy, bounding streams, etc. are similar, and SRS PAs simulate subsurface groundwater flow using the GSA model. Each PA modifies the GSA model to provide increased resolution within the localized area being simulated.
Figure 13: GSA Model, Including Water Table Contours and Stream Traces, From Which Local SRS Facility Flow Models Were Extracted (Figure 4.4109 from the 2020 SDF PA)
The Aquifer Transport Model is such a local model for the SDF and part of the 2020 SDF PA. It was developed using the fully threedimensional GSA model and encompasses the Z-Area.
3.3.1 Regional Model: The GSA Groundwater Flow Model The NRC staff had previously documented its review of the 2016 and 2018 GSA models in the 2016 and 2018 GSA Models TRR; however, the emphasis of the review pertained to SRS Tank Farm Facilities, specifically the areas around the FTF and HTF. This TRR relies on the 2016 and 2018 GSA Models TRR for general conclusions made on technical aspects of the 2016 and 2018 GSA models, but not on conclusions specific to the SDF area.
3.3.1.1 Hydrostratigraphy, Model Domain, and Construction The PORFLOW model domain extends from the Fourmile Branch on the south, Upper Three Runs on the north, F-Area on the west, McQueens Branch on the east, and from the ground surface to the bottom of the Gordon Aquifer. The three-dimensional mesh includes 108 elements along the east-west axis, 77 elements along the north-south axis, and 20 elements with depth, resulting in over 175,000 nodes. All mesh elements are approximately 61 m x 61 m (200 ft x 200 ft).
Model layers are truncated where appropriate (i.e., where actual hydrostratigraphic unit pinches out). The hydrostratigraphic units of the site are represented by various number of GSA model layers: the Gordon Aquifer Unit is represented by two model layers; the Gordon Confining Unit is represented by two model layers, the UTRA-LAZ has one to five model layers, the TCCZ has two model layers, the UTRA-UAZ can have up to 10 model layers and is represented by the Transmissive Zone (TZ) and by the AAA zone above the TZ (note: the letters AAA are not an abbreviation for a specific name). The total thickness of the TZ is approximately 10 m (30 ft) which correlates with a more transmissive member of the Dry Branch Formation. The TZ is thought to be the primary pathway for contaminant transport in the UTRA-UAZ. The GSA hydrostratigraphic model layers can be identified in Figure 14.
Figure 14: GSA Cross Sections at the SDF*: Applied Horizontal Saturated Hydraulic Conductivities (Table 4.484 from the 2020 SDF PA)
- The upper left view shows the SDF area; the lower left cross section shows the Upper Three Runs to the east (left) and McQueens Branch to the west (right); the upper right shows a cross section running through the Z-Area with the Upper Three Runs to the north on top of the Gordon Aquifer with its two model layers of relatively high hydraulic conductivity; the lower right combines the two cross sections.
3.3.1.2 Boundary Conditions and Material Properties Groundwater from the UTRA unit is assumed to discharge equally from each side of Upper Three Runs, Fourmile Branch and McQueens Branch providing natural, no-flow boundary conditions. Hydraulic head values from a contour map of measured water elevations are prescribed for the western boundary. For the deeper Gordon Aquifer Unit, a no-flow boundary condition is also created on the northern boundary by assuming equal discharge from both sides of Upper Three Runs. Prescribed hydraulic head values are used for the west, south and east faces of the model within the Gordon Aquifer since there are no other natural boundary conditions available (Figure 15). A combined recharge/drain boundary condition is applied over the entire top surface of the model. A general head boundary condition of 55 m (180 ft) was applied to the base of the 2016 and 2018 GSA models, differing slightly from the 59 m (195 ft) used in the 2004 GSA model (DOE document SRNL-STI201700008).
Figure 15: Four Types of Boundary Conditions Using the 2004 GSA Model as an Example (Figure 22 from WSRC-TR200400106)
Initial hydraulic conductivity values in the model are based directly on a large characterization database (i.e., the GSA Database Model starting from 1980s). The vertical hydraulic conductivity of the Gordon Confining Unit in the 2016 GSA model is 2.3 x 105 m/d (7.5 x 105 ft/d) because it better matched previous characterization data and regional groundwater flow modeling efforts and to achieve better agreement with well water-level calibration targets (DOE document SRNL-STI201700008). The vertical hydraulic conductivity of the Gordon Confining Unit in the 2018 GSA model was changed back to a value of 3 x 106 m/d (1 x 105 ft/d). Due to
this less permeable confining layer, groundwater travel preferentially occurred to a greater degree in the shallow UTRA (DOE document SRNL-STI201800643).
3.3.1.3 Recharge and Stream Baseflow The 2016 and 2018 GSA models assume a uniform recharge rate of 380 mm/yr (15 in/yr) (DOE document SRNL-STI201700008) for the GSA excluding the discharge areas and the seepage areas/basins which were assigned a recharge of 457 to 483 mm/yr (18 to 19 in/yr). If the discharge areas where water is leaving the system is included in an average recharge rate over the entire model domain, this average approximates 300 mm/yr (12 in/yr) (DOE document SRNL-STI201700008). The range of uncertainty for the uniform recharge rate is estimated at 254 to 406 mm/yr (10 to 16 in/yr) (DOE document WSRC-TR9900248). Although the DOE indicated that local recharge across the GSA may vary with elevation, topography, and vegetation cover, and stated that large-scale variation in topography could increase recharge in the center of the model and decrease recharge elsewhere, the DOE also indicated that no compelling evidence for significant, large-scale variations in recharge was found. Additional assumptions include GSA engineered covers returning to background infiltration rates within a 2-to-4-year period (DOE document SRNL-STI201500351) unless they are overlain by a geomembrane and at least 1.8 m (6 ft) of soil.
The DOE used parameter estimation software (PEST) to automate calibration of the GSA model. During calibration, the uniform recharge rate was not treated as a calibration parameter to reduce PEST runtime and sidestep issues associated with mathematical non-uniqueness (DOE document SRNL-STI201700008). That is, recharge was treated as a constant to allow PEST to fit hydraulic conductivity to water well-level calibration targets. The DOE justified this assumption stating that recharge and hydraulic conductivity cannot both be independently estimated unless stream baseflow measurements are included in the set of calibration targets along with well water-level elevations. The DOE also indicated that this approach allows stream baseflow data to be used for model validation (DOE document SRNL-STI201700008).
In the DOE response to NRC RAI Comment FF2 in the DOE document SRR-CWDA20210072, the DOE described the uncertainty associated with the stream baseflow data acquired for the GSA model extent. Due to the uncertainty, overall flowrates in the GSA flow model are controlled by the recharge rate imposed over the top surface, 38 cm/yr (15 in/yr) away from discharge areas (DOE document SRNL-STI201700008), so that overall model flow rates are directly correlated to uncertainty in the assumed recharge rate. The uncertainty in the baseflow data is due to the difficulty in separating the groundwater baseflow discharge component from the surface water component entering the stream. The latter component is closely tied to the duration and intensity of precipitation in the area and can vary significantly from month-to-month and year-to-year. Additional sources of uncertainty include: (1) the locations of the streamflow measuring stations along Upper Three Runs and Fourmile Branch do not directly correspond with the GSA flow model domain boundaries; (2) baseflow estimates based on these measurements do not align with the baseflow estimates from the GSA flow model; (3) baseflow to a stream such as the Upper Three Runs is assumed to be equal on both sides of the stream although the actual baseflow rate will vary locally based on features such as topography, vegetation, pavement, and watershed surface area; and (4) anthropogenic water discharges; and v. streamflow estimates calibrated to stream depth measurements tend to become biased over time due to changing stream channel morphology.
In the DOE response to NRC RAI Comment FF3 in the DOE document SRR-CWDA202100072, the DOE provided the basis for a sensitivity study that investigates the potential significance of non-uniform present-day recharge on GSA flow model development.
Specifically, recharge rates were varied for those two areas within the SDF that will have closure caps. One sensitivity case increases the recharge rate by 10 cm/yr (4 in/yr) while the other sensitivity case decreases the rate by 10 cm/yr (4 in/yr). Using the new recharge boundary conditions, both flow models were calibrated to present-day hydraulic head calibration targets and matching hydraulic conductivities. The higher local recharge rates cause concentration peaks to decrease by approximately 15% and groundwater pathlines in the northern sector to diverge while lower local recharge rates cause concentration peaks to increase by approximately 15% and flow lines to contract.
3.3.1.4 Calibration Targets Uncertainty in the calibration targets was considered by weighting the hydraulic head residuals.
Greater weight was given to calibration targets estimated with more water-level elevation data and less variability over time (DOE document SRNL-STI201700008). Geographic weights were also applied to account for clustering or sparsity of water well calibration targets across the GSA (DOE document SRNL-STI201500351, Appendix D).
The DOE focused updated GSA groundwater model calibration on reproducing steady-state hydraulic heads/well water-level elevations. Wells having historical GSA water-level data became common after 1987 and data associated with time periods characterized by what DOE described as anomalously low or high precipitation totals were intentionally avoided, as were time periods when artificial pumping or recharge occurred, so that a Base Period was identified during which the running average of precipitation events remained relatively constant. For the 2018 GSA model, this Base Period extends from January 1, 2004, through April 1, 2018.
New Z-Area well data was acquired during the period from 2014 to 2018 and available for use as 2018 GSA calibration targets (DOE document SRNL-STI201800336) including data from 19 wells at the SDF. The number of Z-Area wells screened in the UTRA-UAZ increased from three to six, and those screened in the UTRA-LAZ increased from seven to eleven (DOE document SRNL-STI201800643). For the entire 2018 GSA model domain, new mean well water-level calibration targets were defined for 339 UTRA-UAZ wells, 294 UTRA-LAZ wells, 81 wells in the Gordon Aquifer, and 6 TCCZ wells, including 2 wells screened below that unit.
3.3.1.5 Model Calibration and Validation Saturated hydraulic conductivity values based on model calibration performed using a PEST calibration code were assigned to each subsurface hydrostratigraphic unit (DOE document SRNLSTI201800643). Local modeling of the Z-Area used these hydraulic conductivity values since they were calibrated to local well data. Relative horizontal hydraulic conductivity for the model layers is shown in Figure 14. Table 4 summarizes the vertical and horizontal saturated hydraulic conductivity values obtained. Four different optimization/calibration cases were assessed for goodness-of-fit during the 2016 GSA model calibration effort, including PEST optimization. These consisted of layer-cake vs. heterogeneous hydraulic conductivity fields and use of unweighted vs. weighted hydraulic head residuals. Altogether, 53 PEST simulations were conducted, and all optimization cases run with PEST included global multipliers to the initial UTRA-LAZ, TCCZ, TZ, and AAA zone hydraulic conductivity fields (DOE document SRNL-STI201700008). Elliptical or polygonal regions were added locally to modeled H-and Z-Areas to improve model calibration. Although the Z-Area has only a few, isolated calibration targets, an
improved match to Z-Area well water-level calibration targets was attained after additional PEST simulations were conducted to address specific calibration issues. The layer-cake, weighted model (PEST.51/GSA2016.LW) was recommended as the baseline groundwater simulation for PAs and related analyses (DOE document SRNL-STI201700008) because fewer assumptions and input parameters were required, and it had a better fit to well water-level calibration targets.
The updated GSA_2018.LW model was calibration using the calibrated 2016 GSA model. A vertical hydraulic conductivity value for the Gordon Confining Unit of 3 x 106 m/d (1.0 x 105 ft/d) resulted in a horizontal hydraulic conductivity value of 4.51 m/d (14.8 ft/d) for the TZ zone in Z-Area, higher than the UTRA-LAZ hydraulic conductivity value, thereby improving agreement with the hydrogeologic conceptual model for the Z-Area which conceptualizes a stronger lateral flow component above the TCCZ.
Table 4: Summary of Applied Vertical and Horizontal Saturated Hydraulic Conductivities (Table 4.484 from the 2020 SDF PA)
Log10Kv Vertical Saturated Hydraulic Conductivity Log10Kh Horizontal Saturated Hydraulic Conductivity (ft/day)
(ft/day)
(cm/s)
(ft/day)
(cm/s)
AAA
-0.866 1.36E01 4.8E05 0.616 4.13E+00 1.5E03 UAZ
-0.308 4.92E01 1.7E04 1.17 1.48E+01 5.2E03 TCCZ
-2.70 1.99E03 7.0E07
-1.23 5.96E02 2.1E05 LAZ
-0.741 1.82E01 6.4E05 0.736 5.44E+00 1.9E03 GCU
-5 1.00E05 3.5E09
-4 1.00E04 3.5E08 GAU
-0.420 3.8E01 1.3E04 1.58 3.8E+01 1.3E02 The DOE evaluated the impact of the updated 2016 and 2018 GSA flow models on the results of PAs for FTF, HTF, and SDF (DOE documents SRNL-STI201700445; SRR-CWDA201700068). In SRNL-STI201700445, the DOE noted that slower groundwater velocity in eastern Z-Area results in higher I129 concentrations at the 100-m (328-ft) boundary and Upper Three Runs (UTR) seepline.
Simulated baseflow rates from a calibrated model that match the measured baseflow rates reduce the chance of a non-unique solution. The original GSA model validation exercise was to include assembling and evaluating the completeness, representativeness, and accuracy of previously used baseflow data and more recent baseflow data and then to develop optimization weighting factors for stream baseflow to be used to validate the model (DOE document SRNL-STI201600261). However, in the 2019 NRC SDF OOV Report (ML19143A084), the NRC staff was not aware of any attempt to revise stream baseflow estimates since the estimates published in 1997 and apparently no monthly streamflow data is currently being measured in the GSA.
The period of record used to develop baseflow estimates is 1973 to 1995. However, the DOE indicated that these stream baseflows may be higher than normal because rainfall and water levels were higher on average during the 1973-1995 period compared to the 2004-2014 period (DOE document SRNL-STI201700008). In SRNL-STI201700008, the DOE provided the recharge and stream baseline estimates shown in Table 5.
Table 5: Recharge and stream baseflow estimates (modified Table 21 from SRNL-STI201700008)
Flow parameter Estimate Recharge 38 cm/yr (15 in/yr)
Baseflow from GSA model domain:
Upper Three Runs and tributaries excluding McQueens Branch 0.515 m³/s (18.2 ft3/s)
Fourmile Branch and tributaries 0.074 m³/s (2.6 ft3/s)
McQueens Branch 0.042 m³/s (1.5 ft3/s)
Crouch Branch 0.051 m³/s (1.8 ft3/s) 3.3.2 Local Model: The SDF Aquifer Transport Model Using the full GSA model to simulate flow and transport within a small portion of the model is computationally inefficient. Therefore, the DOE developed a more local SDF Aquifer Transport Model (ATM) using many aspects of the fully calibrated threedimensional GSA model.
Specifically, the ATM was constructed by extracting a smaller grid from the GSA model using only those model nodes relevant to the SDF and then refining the GSA model to generate localized flow fields and finally implementing the actual contaminant transport modeling. The ATM was used to simulate saturated zone transport. Contaminant fluxes from the Vadose Zone Transport Model and flow modeling from the GSA model created contaminant concentration estimates at points of assessment, such as 100m (328 ft) boundary wells. These concentration values are then used as input to the Dose and Exposure Pathways Model to estimate doses to future human receptors.
3.3.2.1 Model Domain and Grid The original SDFspecific portion of the GSA model used for ATM had a discretization too large for the smaller model domain. Rather than a separate flow model requiring its own boundary conditions and properties, in the 2020 SDF PA the DOE generated a velocity field directly from the coarser-scale GSA model by subdividing the 61 m x 61 m (200 ft x 200 ft) GSA grid cells. A massconserving linear interpolation scheme was then used to assign velocities to the refined 7.6 m x 7.6 m (25 ft x 25 ft) horizontal mesh of the ATM thereby reducing the chance of excessive numerical dispersion (DOE document SRNLSTI201800012). The local velocity field included the entire vertical extent of the GSA model (i.e., the bottom of the Gordon Aquifer Unit) within the horizontal confines of the ATM domain (Figure 16).
Figure 16: Model Domain of the ATM Including the 25 Feet by 25 Feet Grid Discretization (Figure 4.4114 from the 2020 SDF PA)
Contaminant fluxes from the Vadose Zone Transport Model beneath each disposal structure were not inventory specific. The DOE needed to scale the fluxes to reflect the inventory of each disposal structure. Any grid cell with its centroid lying within the footprint of a disposal structure was defined as a source node. The scaled contaminant fluxes were then divided proportionally based on the source nodes and applied to these nodes. Nodes adjoining on any side or corner with other nodes inside the 1m (3.3-ft) boundary were defined as 1m (3.3-ft) perimeter cells.
Nodes adjoining on any side or corner with other nodes inside the 100m (328-ft) boundary were defined as 100m (328-ft) perimeter cells. The 100m (328-ft) boundary was divided into eight sectors labeled A though H for evaluation purposes (see Figure 16 for the one-and hundred-meter boundaries and the eight sectors).
3.3.2.2 Dispersion, Recharge, and Contaminant Flux
Soil and hydraulic properties, soil Kd values, refined local velocity field (as discussed above),
contaminant fluxes from the Vadose Zone Transport Model, and dispersivities used to define plume spreading are part of the ATM.
As described in the 2020 SDF PA, the ATM included dispersivity values, used to define plume spreading, based on mechanical dispersion coefficients. Longitudinal and transverse dispersivities are 10% and 1% of a nominal 100m (328-ft) plume travel distance, or 10 m (33 ft) and 1 m (3.3 ft), respectively, and determine horizontal plume spreading via mechanical dispersion. Vertical plume spreading via mechanical dispersion is determined based on assumed vertical longitudinal and transverse dispersivities of 1 m (3.3 ft) and 0.1 m (0.3 ft), or 1/10th of the horizontal dispersivities, respectively.
Flow exiting the Vadose Zone Flow Model at the water table is not used to define local recharge beneath the SDF within the ATM even though the planned SDF will have an engineered surface cover that is intended to reduce the amount of groundwater considerably. Rather, the average recharge rate from the GSA model is used as flow rate input to the upper boundary of the ATM (i.e., the water table). However, despite the flow rate disconnect between the output from the Vadose Zone Flow Model and the input to the ATM, the ATM used the contaminant flux of the Vadose Zone Transport Model as input. Flux contributions from individual disposal structures were assigned to the ATM grid by uniformly distributing the flux to those water table cells with centroids lying within the footprint of the disposal structure.
3.3.2.3 Results As an intermediate result, the DOE simulated the centerline streamline traces based on the refined model grid from each disposal structure. These streamlines represent the center of mass along a progressing plume and not the threedimensionality of the contaminant transport.
Tracer plumes from the ATM based on the steadystate release of a hypothetical tracer can better capture the threedimensionality of the contaminant transport. The DOE results for SDS 6 and SDS 7 showed significant spreading, which the DOE attributed to a groundwater divide that runs between the disposal structures. Figure 13 illustrates this groundwater divide by showing stream traces from many of the disposal structures ending up near the Upper Three Runs, while the rest are carried to the McQueens Branch. This means that contaminants released north of the groundwater divide are transported in a generally northward direction, while contaminants released south of the groundwater divide are transported in a generally eastward direction.
Disposal structure plumes can be transported in both directions, and be subject to spreading, if the disposal structure source is near the groundwater divide.
Breakthrough curves from the ATM were also presented in the 2020 SDF PA and represent the concentration of a hypothetical tracer pulse at the 100m (328-ft) SDF boundary. The peaks in these curves indicate how long it takes for the center of mass of a plume to reach the boundary from the time it is released into the saturated zone beneath the respective disposal structures.
Streamline traces generated in the ATM were also used to estimate the breakthrough times by exporting coordinate location and flow rates from each streamline trace and calculating the average flow rate to the 100-m (328-ft) boundary. Comparing the average flow rates between tracer pulses/breakthrough curves and streamline traces show differences in flow rates between the two approaches and between the disposal structures themselves. These differences were evaluated in the benchmarking analysis for the SDF GoldSim Model (Section 5.6 in the 2020 SDF PA).
3.3.2.4 Uncertainty: Probabilistic Analyses The 2020 SDF PA performed probabilistic sensitivity analyses and provided both PRCCs and SRRCs. Probabilistic sensitivity analyses help identify parameter distributions that have the greatest influence on the PA results.
The PRCCs for total dose to a MOP at any sector are shown in Figure 12. The saturated zone width for SDS 7 (labeled SDU7_SatWidth in Figure 12) is a parameter of some importance within the 1,000-year timeframe and slightly more importance during the 10,000-year timeframe.
The saturated zone width is essentially the same as the bottom area of the disposal structure, and for the analysis, a uniform distribution ranging from 80% to 120% of the cylindrical disposal structure diameters was sampled. The stepwise regression analysis calculated both the saturated zone widths for SDS 7 and for SDS 6 as having higher coefficients as can be seen in Table 5.78 in the 2020 SDF PA; however, these saturated zone widths were not one of the four or five variables with the highest coefficients which effectively dominate any influence over the dependent variable at the time analyzed.
3.3.2.5 Uncertainty: Deterministic Sensitivity Analyses Deterministic sensitivity analyses also examined the sensitivity of saturated thicknesses and saturated hydraulic conductivity (using flowrate multipliers) using climate change scenarios.
Other parameters such as cementitious degradation rate and initial saltstone hydraulic conductivity were also modified to capture more extreme flow conditions in addition to different infiltration rates. As indicated in Section 3.2 of this TRR, the results of the analyses showed that over the compliance period, the peak dose for the wetter climate case was approximately 3.7 times higher than the peak dose for the Compliance Case, and the peak dose for the Compliance Case was approximately 3.5 times higher than the peak dose for the drier climate case.
Current DOE plans for the SDF include an effective engineered surface cover that is expected to considerably reduce the groundwater recharge below the cover and disposal structures.
However, as previously discussed, the recharge rate to the ATM does not differ from the recharge rate applied in the GSA model (380 mm/y, or 15 in/yr). To evaluate the potential impacts from this reduced recharge, two sensitivity cases using a modified GSA model were run (Section 5.8.8.4 in the 2020 SDF PA) without recalibrating. One reduces the groundwater recharge rate to the area beneath the SDF closure caps to the longterm estimate for infiltration through the cover under Compliance Case conditions (0.13 mm/yr, or 0.51 in/yr). The second sensitivity case is the same as the first but models any water that drains off the sides of the closure cap to the surrounding perimeter as infiltrating back into the groundwater. The DOE presented figures with streamline traces for the Compliance Case and the two sensitivity cases.
For the sensitivity cases, the streamline traces from the disposal structures traveled in a more northwesterly direction beneath the cover. Tenyear markers in the figures indicated that the reduced recharge rates had a small influence on the groundwater flow rates; flow rates were slightly slower under the reduced recharge conditions. The contaminant fluxes from the Vadose Zone Transport Model were input into these modified groundwater flow models to estimate groundwater concentrations. Reduced recharge rates increased dose estimates more for the performance period than for the compliance period as can be seen in Table 6. Besides reduced dilution, the DOE attributed the increase to slight changes in flow directions of the plumes and therefore to more overlapping of the plumes.
Table 6: Comparison of the 100-Meter MOP Peak Groundwater Pathways Doses Based on Reduced Groundwater Recharge Rates (Table 5.832 from the 2020 SDF PA)
Modeling Case Compliance Period (0 to 1,000 Years)
Performance Period (0 to 10,000 Years)
Peak Dose (mrem/yr)
Year of Peak Peak Dose (mrem/yr)
Year of Peak Compliance Case (Natural Recharge = 380 mm/yr) 9.4E03 1,000 1.2 10,000 Cap Recharge = 0.13 mm/yr 1.3E02 1,000 1.8 10,000 Cap Recharge = 0.13 mm/yr + Drainage 1.2E02 1,000 1.8 10,000 The DOE stated in the 2020 SDF PA that it was not possible to validate the future flow conditions simulated within these cases until the closure cap has been put in place. Also, the DOE discussed a decreasing water table elevation due to the closure cap. This would increase the transport distance in the unsaturated zone and likely lower the dose rates.
Although the DOE stated that soft zones in the area of SRS will likely play an insignificant role in the overall PA results, uncertainty regarding the presence of potential soft zones remains. Soft zones consist of clastic material which may decrease transport time. The DOE developed sensitivity cases involving fast flow paths through groundwater to observe the influence they may have on performance.
The first soft zone sensitivity case assumes that the soft zone thickness is one-tenth that of the saturated zone and is fully saturated and continuous. The second sensitivity case is similar to the first but such that the radionuclide concentrations are fully mixed (i.e., source zone thickness and softzone thickness are vertically equal). Full vertical mixing has not been attained by the time mass reaches the 100meter boundary for the compliance case, so the third sensitivity case modified the compliance case such that the radionuclide concentrations are fully mixed to better evaluate the influence of mixing since full vertical mixing is not attained in the compliance case by the time contaminants reach the 100m (328-ft) boundary. A fourth simulation lowered the soft zone transverse vertical dispersivity by a factor of 100.
For the first soft zone case, a continuous fast zone reduced the dose because it results in a dilution effect. That is, the degree of vertical mixing due to dispersion is a function of distance traveled due to advection resulting in full vertical mixing. For the second case, dose results were nearly identical to the first, confirming that the former simulation results represent nearly complete vertical mixing. Results from the third sensitivity case also are very similar to the first two due to total vertical mixing. Figures in the 2020 SDF PA comparing the fourth sensitivity case results with that of the compliance case showed that the consistent scaling of geometry, flow rates, and vertical dispersivities produced similar results.
The colloid transport sensitivity cases assumed that radionuclides would contact and sorb to naturally occurring microscopic clay colloids that remain suspended in the water, resulting in much faster transport than retarded radionuclides in equilibrium with immobile solids. These sensitivity cases were performed because the Central Scenario does not explicitly address the potential for colloid formation or transport. For these sensitivity cases, the DOE assumed that the colloids migrate at the same rate as a nonsorbing species (tracer) and are subject to the same degree of dispersion.
A colloid transport sensitivity case involved modifying the Compliance Case by replacing Kd value for sand with the effective equilibrium coefficients (as defined in Eq. 5.89 of the 2020
SDF PA) based on a colloid concentration of 3 x 108 g/mL. Further sensitivity cases involved increasing the colloid concentration to 3 x 104 g/mL. The figures in Section 5.8.9.6 of the 2020 SDF PA show the influence of colloid facilitated transport on the total dose for these sensitivity cases are negligible, although the influence on Kds for highly sorbing radionuclides becomes even greater. The DOE explained the lack of sensitivity as being likely due to the lack of stronger sorbing radionuclides, such as americium and plutonium, over a period of 40,000 years due to the effectiveness of the engineered barriers.
As discussed in Section 3.2 of this TRR, a sensitivity case evaluated general Kd sensitivities including the Kd values of the native soils by decreasing all the Kds by a factor of 100. Overall, the impact of these changes was minor since I129 and Tc99 dominate dose, and both have relatively small Kd already.
4.0 NRC Staff Evaluation As described in Section 1.0 in this TRR, the NRC staff reviewed aspects of the 2020 SDF PA and supporting references related to SDF hydrogeology, groundwater monitoring at the SRS, and the models the DOE used to simulate flow and transport in the unsaturated zone and saturated zone. During that review, the NRC staff determined that MA 7 would be more accurately titled by including the term flow within the title because groundwater flow is a risk-significant process affecting performance.
Recommendation HFFM-01 Therefore, the NRC staff recommends that MA 7 (Subsurface Transport), be relabeled as MA 7 (Subsurface Flow and Transport) because flow is a risk-significant factor that requires monitoring.
4.1 Groundwater Monitoring The NRC staff will monitor groundwater data for the duration of NRC monitoring at the SDF under MF 8.02. Technical topics discussed in this section are related to MF 8.02. MF 8.02 is an open, periodic monitoring factor (i.e., monitoring factor related to data that the NRC staff expects to review on a periodic basis) and addresses identifying early releases of radionuclides from saltstone whereby the NRC staff reviews groundwater data from the groundwater monitoring system for any indication of unintentional release of contaminants. As stated in Section 2 of this TRR, much of the following information concerning groundwater monitoring prior to 2018 relied on information from the 2018 Groundwater Monitoring TRR.
The NRC staff reviewed information related to the DOEs ability to identify saltstone contaminants in the groundwater and ability to track contaminant plume movements under MF 8.03. Technical topics discussed in this section are related to MF 8.03. The NRC expects to close MF 8.03 under POs §61.41 and §61.42 when the NRC staff determines that the groundwater monitoring system in the Z-Area can: (1) identify saltstone contaminants in the groundwater in the SDF at no more than 150 ft (46 m) from a disposal structure; and (2) track the movements of the groundwater plume (e.g., know the horizontal and vertical extent of the plume; be able to follow the approximate path of the peak of the plume).
In the review of the 2020 SDF PA, the NRC staff found that the current groundwater monitoring system for the Z-Area is not adequate because of inadequate monitoring wells in the UTRA-UAZ, as described in further detail later in this section. That finding is consistent with previous NRC reviews. As documented in the 2015 NRC SDF OOV Report (ML15041A562), the NRC
staff pointed out that there were relatively few wells screened above the TCCZ in comparison to the wells screened below the TCCZ. All planned disposal structures except for SDS 1, SDS 4, and possibly SDS 2A and SDS 10, (i.e., SDS 3, SDS 5, SDS 6, SDS 7, SDS 8, SDS 9, SDS 11, SDS 12) cannot be adequately monitored for contaminants in the UTRA-UAZ because no monitoring wells downgradient from those disposal structures are screened in this hydrogeological unit.
During the NRC review of the 2020 SDF PA, the DOE response most directly linked to the concern of insufficient UTRA-UAZ monitoring wells is the DOE response to FF4 C in the DOE document SRR-CWDA202100072. Table FF4.1 in that document (i.e., Table 7 below) presents the screened zones for Wells ZBG 17D, ZBG 18D, and ZBG 19D, the associated estimated elevation of the top of the TCCZ, and the latest water elevation for the respective wells based on the DOE documents K-ESR-Z00005 and SRNS-TR202100362. The DOE pointed out that the water elevations for the three wells are all lower than the top of the TCCZ.
Table 7: Information on Monitoring Wells ZBG 17D, ZBG 18D, and ZBG 19D (modified Table FF4.1 from SRR-CWDA202100072)
Well Number Screen Zone Top (ft Above MSL)
Screen Zone Bottom (ft Above MSL)
Nearest CPT to Well TCCZ Top (ft Above MSL)
TCCZ Thickness (Range) (ft)
Elevation of the Potentiometric Surface*
(March 2021)
(ft Above MSL)
ZBG017D 222.9 212.9 Z-SDU6-C22 237 7
220.09 ZBG018D 215.1 205.1 Z-SDU6-C18 235 18 223.12 ZBG019D 218.3 208.3 Z-SDU6-C04 236 7
222.67
- The original Table FF4.1 from SRR-CWDA202100072 reported these elevations as Water Table Elevations.
Monitoring Wells ZBG 17D, ZBG 18D, and ZBG 19D are all screened in the UTRA-LAZ and the water measured in that unit actually represents a potentiometric surface and not the free water surface or water table. The DOE document K-ESR-Z00005 is dated April 2012. More recent information is available in the DOE document SRR-CWDA202100065, which presented the new 2021 GSA model with the latest interpretation of the elevation for the top of the TCCZ.
Figure 3 in that document showed the top of the TCCZ in feet and outlines for SDS 6, SDS 7, and SDS 8 are visible. Based on Figure 3, it appears that the information in the DOE document K-ESR-Z00005 is outdated and the TCCZ top elevations for Wells ZBG 17D and 18D are below 67 m-amsl (220 ft-amsl), while the top of TCCZ for Well ZBG 19D would be above the water elevation given in the table. Comparing Figure 9 from DOE document SRR-CWDA202100065, which shows simulated water table levels and UTRA-UAZ hydraulic heads from the 2021 GSA model, with Figure 3 from that same document, it appears that the soil below SDS 2B, SDS 3A, SDS 3B, SDS 5A, SDS 5B, SDS 7, SDS 8, SDS 9, SDS 11, and SDS 12 has sufficient water for a water table to be present in the UTRA-UAZ and therefore sufficient water to be monitored. In addition, the modeling results from the 2020 SDF PA show lengthier lateral groundwater transport occurring in the UTRA-UAZ from most of the disposal structures (i.e., the distance traveled in the UTRA-UAZ is not insignificant for most disposal structures).
Figure 17 (Figure CC7.6 in the DOE document SRR-CWDA202100072) presents light blue lines emanating from the disposal structures representing transport in the UTRA-UAZ.
Figure 17: Particle Tracking in the ATM for the Disposal Structures Approximately Showing Transport in the UTRA-UAZ [light blue], TCCZ [orange],
and the UTRA-LAZ [violet] (Figure CC7.6 from SRR-CWDA202100072)
Table CC7.3 in the DOE document SRR-CWDA202100072 compiled similar information and grouped transport from the disposal structures into three groups based on speed: slower travel times due to predominate transport within the UTRA-LAZ, intermediate speeds due to transport within the UTRA-UAZ and UTRA-LAZ, and a faster group due to predominant transport within the UTRA-UAZ. Based on the table, dose estimates are all influenced by flow and transport within the UTRA-UAZ with the possible exception of the SDS 1, SDS 2A, SDS2B, SDS 4, and SDS 10. The DOE response to CC8 in the DOE document SRR-CWDA202100072 confirms this with the statement that, Viewing the UAZ layer directly above the TCCZ in layout view shows that the releases from SDS 7, SDS 8, SDS11, and SDS 12 are all loaded, in their entirety, within the UAZ of the Aquifer Transport Model (Figure CC-8.3). With regard to SDS 1, SDS 2A, SDS 2B, SDS 4, and SDS10, the NRC staff also concluded that these disposal structures may also be strongly influenced by flow and transport within the UTRA-UAZ, based on information and insights obtained from the SDS4 plume data.
The DOE response for CC5 in SRR-CWDA202100072 explained the significance of the color scheme for Figure 4.4113 in the 2020 SDF PA, which shows discharge points for the Gordon Aquifer and the UTRA into the Upper Three Runs and the McQueens Branch. In Figure 4.4113, (see Figure 18 below) the orange squares represent stream discharge points from the UTRA-
LAZ, the green squares represent stream discharge points from the UTRA-UAZ, and the blue diamonds represent stream discharge points from the Gordon Aquifer.
Figure 18: Streamlines and Discharge Points for SDF-Specific Portion of the GSA Model (Figure 4.4113 from 2020 SDF PA)
The green squares are visible between the SDF and the McQueens Branch. Figure 18 shows that none of the streamlines emanating from SDS 1, SDS 2A, SDS 2B, SDS 4, or SDS 10 and traveling toward the McQueens Branch area are modeled as discharging in these green square discharge points. Although this may be plausible, the discharge points for UTRA-UAZ indicate
not all of the UTRA-UAZ can be dry between the Z-Area and McQueens Branch. That, despite Well ZBG 16D being dry, some groundwater is traveling in this upper zone for most of the distance between the SDF and the McQueens Branch and that is unlikely that all contaminants from SDS 1, SDS 2A, SDS 2B, SDS 4, and SDS 10 will travel most of the path within the UTRA-LAZ. In addition, in the DOE document SRNS-RP201500902, the DOE presented the results of the Z-Area groundwater characterization data showing that the SDS4 plume had moved beyond the Z-Area boundaries while traveling in the UTRA-UAZ.
The information discussed above indicates that water from the UTRA-UAZ is present near most of the disposal structures and that modeled radionuclide travel occurs in the UTRA-UAZ.
Information from the DOE document SRR-CWDA202100065 indicates that the UTRA-UAZ does exist near most of the disposal structures. Evidence from the SDS 4 plume provides evidence that significant lateral transport within the UTRA-UAZ is possible. Unfortunately, due to the lack of monitoring wells screened in the UTRA-UAZ, any present-day leaks from most of the disposal structures would currently not be detected until contaminants had traveled through the TCCZ to the UTRA-LAZ, where most monitoring wells are located. It is for these reasons, that the NRC staff has concluded that the current groundwater monitoring system is not adequate.
Because the NRC staff finds the current SDF groundwater monitoring system not adequate, the NRC staff determines that the DOE cannot identify new plumes in a timely fashion nor adequately monitor either the current SDS 4 plume or potential future plumes in the UTRA-UAZ.
The current groundwater monitoring system cannot adequately monitor the SDS 4 plume as it may exist today because, based on groundwater contours shown in the DOE document SRNS-TRR202100966 and saturated zone modeling results, no monitoring wells appear to exist downgradient of Wells ZBG 2C and ZBG 2D. New plumes cannot be identified in a timely fashion due to the lack of UTRA-UAZ groundwater monitoring wells in the SDF for all current and planned disposal structure except for SDS 1, which has groundwater monitoring wells in the UTRA-UAZ nearby, and for SDS 2A and SDS 10. The NRC staff based that determination on the comparison between Figure 3 and Figure 9 from the DOE document SRR-CWDA2022100065, which indicated that the UTRA-UAZ may not have sufficient water at those locations for it to transport contaminants. The NRC staff has determined that based on the most plausible conceptual model of flow and transport at the SDF, there needs to be groundwater monitoring wells that are screened in the UTRA-UAZ, and wells screened in the UTRA-LAZ, when monitoring disposal structures located in areas where the UTRA-UAZ is present.
Evidence in the 2015 NRC SDF OOV Report (ML15041A562) indicated that contaminants from SDS 4 have traveled through the unsaturated zone and were then transported in the saturated zone on top of the TCCZ toward Well ZBG 2 and beyond. The results of the 2015 Z-Area groundwater characterization study provided convincing evidence that the plume has traveled past the boundaries of the Z-Area (especially the results of ZDPT 10 and ZDPT 11) and that horizontal transport is much more dominant than vertical transport. Groundwater monitoring reports have provided data showing the pulse of contaminants from the SDS 4 plume for both the UTRA-UAZ and the UTRA-LAZ. From the 2018 Groundwater Monitoring TRR, Wells ZBG 2, ZBG 2D, and ZBG 20D were shown nonvolatile gross beta measurements above 30 pCi/L with the highest measurements recorded so far being 158.0 pCi/L, 132.0 pCi/L, and 51.9 pCi/L, respectively, while the highest concentration values for the groundwater monitoring wells in the UTRA-LAZ showed considerably lower values: Well ZBG 2C with 11.5 pCi/L, Well ZBG 3 with 4.3 pCi/L, Well ZBG 4 with 6.7 pCi/L, and Well ZBG 5 with 3.6 pCi/L. Although the vertical thickness of the TCCZ is typically less than 6 m (20 ft), preferential flow and transport appears to be in the lateral direction. For the NRC staff, the results of the 2015 Z-Area groundwater
characterization study have clearly shown that the TCCZ has the potential to slow the vertical movement of contaminants when compared to the horizontal transport.
The 2021 DOE Z-Area Groundwater Report (DOE document SRNS-TR202100966) did not include the results from the 2015 Z-Area groundwater characterization study when showing the extent of the plume (see Figure 19 below). The dimensions of the plume from the 2021 Groundwater Report were considerably smaller than shown in the 2015 Z-Area Groundwater Report (e.g., the plume length has been shortened to the east and the south). No UTRA-UAZ data on nonvolatile gross beta and Tc-99 exists, so that the delineation of the plumes boundaries to the north is not possible.
Figure 19: Dashed Light Green Line Represents the DOEs Assumed Location for Tc-99 Groundwater Plume (Figure 11 from SRNS-TR202100966)
Figure 7 from the 2017 Z-Area Groundwater Report showed peak concentration values for Well ZBG 2 and Well ZBG 2D having occurred in 2015. As suggested by that report, this may indicate that the peak of the plume has moved on. Knowing the location of the peak of the plume and in what direction it is heading would allow a better evaluation of the potential safety concerns emanating from the plume. Currently the lateral and vertical extent of the SDS 4 plume is not known. In addition, the DOE cannot adequately predict or estimate the future development of the plume (i.e., concentrations over time for different locations).
Current knowledge indicates that the current SDS 4 plume is not being sufficiently monitored to delineate its current hydrogeological extent and depth nor to provide sufficient data on the plumes future direction and concentration. No additional monitoring wells exist downgradient of the monitoring Wells ZBG 2C, ZBG 2D, ZBG 20, and old Well ZBG 2. The NRC staff is not aware of any additional characterization of the extent and depth of the SDS 4 plume having been undertaken since the 2015 characterization study documented in the DOE document SRNS-RP201500902. However, the DOE has made progress with simulating flow and transport of the earlier SDS4 plume and, as stated in the DOE response to FF1 C in the DOE document SRR-CWDA202100072 and in the DOE document SRR-CWDA202100065, results from the new 2021 GSA model with its updated TCCZ stratigraphy near SDS 4 and revised moisture characteristic curves yielded a 2021 GSA particle track that is more consistent with the conceptual model of waste release from Cell G of SDS 4.
The DOE response for FF4 B in the DOE document SRR-CWDA202100072 does not rely on this new information and presented an older conceptual model of potential groundwater flow near the surface (see Figure 20 below), which NRC staff had determined to be much less plausible. Based on the northernmost red flowline originating from Cell G in SDS 4 (the DOE expects Cell G is the source the SDS 4 plume), contaminants from that cell would most likely have bypassed Well ZBG 2, depending on the extent of the plume spread, and never have been detected. The DOE also made the argument in the FF4 response that, groundwater flow in the UAZ from releases on the southern end of SDS 4 would be captured by ZBG020D (blue dot with a water level of 229.55 ft above mean sea level (MSL)) as indicated by the red arrows drawn from SDS 4 to be approximately perpendicular to the UAZ groundwater contours (shown in dotted blue lines). However, based on the mismatch between Cell G and the old Well ZBG 2, it is unclear to the NRC staff if that DOE statement is accurate.
Figure 20: UTRA-UAZ Water Table Levels from the 1st Quarter of 2021 (modified Figure FF4.1 from SRR-CWDA202100072)
The DOE has stated in the past that they do not recognize a significant difference between the UTRA-UAZ and the UTRA-LAZ, such that they should be treated as two separate hydrogeological units (DOE response to RAI Comment FF4 in the DOE document SRR-CWDA202100072). The NRC staff agreed with the DOE that there is hydraulic connectivity across the TCCZ; however, the NRC staff disagreed with the DOE that all water table and piezometric measurements can be used without consideration of whether individual measurements represent the UTRA-UAZ or the UTRA-LAZ. Discussions between the NRC staff and the DOE relating to this topic are documented in the 2016 NRC SDF OOV Report (ML16147A197). The NRC staff recommended that the DOE: (1) create a water table map by using water table measurements from the UTRA-UAZ; and (2) create a potentiometric surface map by using potentiometric measurements from the UTRA-LAZ. With the DOE issuance of the Z-Area Saltstone Disposal Facility Groundwater Monitoring Report for 2019 (DOE document SRNS-TR201900326), water table measurement of the UTRA-UAZ and potentiometric measurements of the UTRA-LAZ are presented separately as are PORFLOW calibration results, thereby providing greater insights into the hydrogeological system of the Z-Area.
Therefore, the NRC staff finds that DOE adequately treats the UTRA-UAZ and the UTRA-LAZ as two separate hydrogeological units.
In RAI Sub-Comment FF1 A (ML21321A087), the NRC described discrepancies between elevation data from the characterization data report for the upper and lower boundaries of the TCCZ and the elevation model input for the ATM. The NRC stated that TCCZ elevation differences may have risk-significant consequences for the velocity and direction of the groundwater flowpaths within the UTRA-UAZ. Consequently, the NRC asked that the DOE provide information showing the risk significance to SDF dose projections of using different upper and lower TCCZ boundary elevations. In the DOE response to RAI Comment FF1 in the DOE document SRR-CWDA202100072, the DOE stated that, [t]he GSA_2021 flow model reflects a revised TCCZ stratigraphy that better agrees with currently available TCCZ elevation data in the southern portion of Z-Area downgradient of SDS 4 (SRR-CWDA-2021-00065, Figures 2 and 3). The DOE document SRR-CWDA202100065 described how an alternative GSA groundwater flow model (the 2021 GSA model) that better represented hydrologic conditions near SDS 4 was created. Using the 2018 GSA model, the elevation and thickness of the TCCZ were modified, particularly the tops of the TCCZ and UTRA-LAZ, near and downgradient of SDS 4 using the latest data available. The NRC reviewed the sections relevant to the 2021 GSA model in the DOE document SRR-CWDA202100065 and finds this model to be acceptable because it reflected the best available information on the extent, thickness, and topography of the TCCZ.
Since the installation of the new background well (ZBG 15D) in 2012, the background concentration values used by the DOE to compare with concentration values obtained from the groundwater monitoring wells within the UTRA-UAZ are considered by the NRC to be appropriate. However, the background concentration values used by the DOE to compare with concentration values obtained from the groundwater monitoring wells within the UTRA-LAZ are not considered to be adequate by the NRC. Both monitoring wells for background values (Wells ZBG 1 and ZBG 15D) are located in the UTRA-UAZ so that samples obtained from the UTRA-LAZ are being compared with background concentration values from the UTRA-UAZ. There are currently no monitoring wells obtaining background values for the UTRA-LAZ, although 11 of the 20 groundwater monitoring wells are monitoring contamination in the UTRA-LAZ.
4.1.1 Associated Monitoring Factors MF 8.02: Groundwater Monitoring; and MF 8.03: Identification and Monitoring of Groundwater Plumes in the Z-Area The NRC staff monitors the DOE SDF groundwater monitoring program. A key aspect of that DOE groundwater monitoring program is the placement of the wells. The NRC staff has concluded that for the current groundwater monitoring plan, the placement of monitoring wells near the disposal structures in the UTRA-LAZ is adequate, but the number of monitoring wells in the UTRA-UAZ is not adequate (i.e., there are insufficient monitoring wells in the UTRA-UAZ near the current and planned disposal structures). Similarly, it is important for the wells used to obtain information regarding the natural groundwater composition (i.e., background wells) to be upgradient of the disposal structures. Currently, the number of upgradient background monitoring wells in the UTRA-LAZ is not adequate. The NRC staff plans to monitor this issue as the revised MF 8.03. Because MF 8.02 is a periodic monitoring factor where the groundwater data will be monitored in perpetuity at the SDF, there is no priority status associated with this monitoring factor. The DOE groundwater monitoring system in the Z-Area will be providing useful information on the hydrogeological system in the Z-Area for the indefinite future.
Recommendation HFFM-02 Therefore, the NRC staff recommends modifying the high-priority MF 8.03 (Identification and Monitoring of Groundwater Plumes in the Z-Area) under MA 8 (Environmental Monitoring) under both PO §61.41 and PO §61.42 because the periodic nature of MF 8.02 cannot address the NRC staff concerns, nor can the periodic MF 8.02 be given a priority status. The NRC staff expects to close MF 8.03 when the NRC staff determines that the groundwater monitoring system in the Z-Area can: (1) identify saltstone contaminants in the groundwater in the SDF at no more than 150 ft [46 m] from a disposal structure; (2) identify saltstone contaminants (as described in criterion (1))
separately in both the UTRA-UAZ and the UTRA-LAZ; (3) obtain background concentration measurements from monitoring wells in the UTRA-UAZ and the UTRA-LAZ upgradient of the SDF; and (4) track the movements of the groundwater plume (e.g., know the horizontal and vertical extent of the plume; be able to follow the approximate path of the peak of the plume).
4.2 Flow and Transport in the Unsaturated Zone Below the Disposal Structures 4.2.1 Vadose Zone Models Used in the 2020 SDF PA The NRC staff finds the general modeling approach used for the vadose zone models to be acceptable for the purpose of estimating flow and transport through the unsaturated native soils because the latest information for hydraulic and physical soil parameter inputs were used and generally acceptable modeling practices were applied. For example, for the Vadose Zone Flow Model, the lower boundary of the model coincides with the water table where pressure head is set to zero, and for the Vadose Zone Transport Model, the bottom boundary condition, or the unsaturated zone/water table boundary, is a zero-concentration boundary condition. The PORFLOW software, used for both the Vadose Zone Flow Model and the Vadose Zone Transport Model, is a widely used flow and transport code for DOE-related projects. Processes such as sorption, precipitation/dissolution reactions, diffusion, and advection were incorporated into the modeling. For example, soil Kd values were modified based on the expected altered chemistry of soils beneath the degrading cementitious material of the disposal structures.
4.2.2 Hydrostratigraphy and Unsaturated Zone Material Properties The NRC staff find the identification and classification of the hydrostratigraphic units near SDF to be acceptable because the classification was consistent with the detailed information provided. For example, the two unsaturated hydrostratigraphic units identified at the SDF, a finergrained (clayey) unit is the upper vadose zone soils and lower vadose zone soils with higher sand content, match soil characteristics described in other sources. Although the technical basis for the demarcation between the vadose zones, an elevation of 80.5 m-amsl (264 ft-amsl) is used as a boundary between the two, is not clear, the upper vadose zone with its higher Kd values is conservatively not represented in the model although compacted lower backfill material properties were developed based on the material properties of the upper vadose zone.
Section 3.2 in this TRR discusses the relative importance during the compliance period of the unsaturated zone thickness. The bottom of the unsaturated zone is the elevation of the water table, which was obtained from the 2018 GSA groundwater model (DOE document SRNLSTI 201800643). Because measured data on water table elevations under the planned disposal structures to the north is scarce, the technical basis for the assumed modeled depth of the unsaturated zone (i.e., from the bottom of the lower mud mat to the water table) is weak. If the
DOE installs groundwater monitoring wells in the UTRA-UAZ as discussed in Section 4.1 of this TRR, then the technical basis for the assumed modeled depth of the unsaturated zone would be strengthened, and the measured water table data obtained from the northern Z-Area could corroborate GSA modeled thicknesses. Because this specific uncertainty associated with the thickness of the unsaturated zone is independent of the potential change in climate and the related changes to the hydrogeological system, the climate change scenarios cannot reduce this uncertainty since other significant parameters (e.g., infiltration rates) are being modified in these runs. The NRC staff will monitor the support for the GSA modeled thicknesses of the unsaturated zone under a newly created monitoring factor entitled Confidence in GSA Modeling Results in MA 7 (see Section 4.3.1.6 of this TRR). Additional monitoring wells in the UTRA-UAZ would provide such support.
The NRC staff find the hydraulic property values of the soils and the moisture characteristic curves for the soils surrounding and below the disposal structures used in the 2020 SDF PA models to be acceptable because the technical basis supporting those values was clear and sufficient (i.e., the data and interpretations were well documented and referenced). The NRC staff also find the assumed homogeneity of the unsaturated zone to be acceptable since sensitivity cases showed hydraulic conductivity value changes within the unsaturated zone did not significantly affect the projected dose.
4.2.3 Sorption Coefficients The NRC staff finds the sorption coefficients the DOE used to model iodine and technetium transport in the subsurface in the 2020 SDF PA to be acceptable for demonstrating compliance with the performance objectives of 10 CFR Part 61 because reasonable alternatives produced either a small increase or a larger decrease in the projected peak dose. The NRC staff agrees with the DOE determination that the data collected from the F-Area North Borrow soil is likely to underestimate iodine sorption in Z-Area because the Z-Area soil contains more sand than the F-Area does. The NRC staff also agrees with the DOE general point that the chemical speciation of iodine in the subsurface is likely to affect iodine sorption; however, the NRC staff finds that the DOE did not demonstrate that the chemical conditions of F-Area soil are applicable to Z-Area. Nonetheless, the NRC staff expects that the uncertainty in the applicability of the chemical conditions is not risk-significant because a DOE sensitivity analysis demonstrated that reasonable changes in the iodine sorption coefficient in site soil only caused a 10% increase in the projected peak dose to a member of the public within 10,000 years of site closure.
The NRC staff finds the leachate-impacted and non-impacted sorption coefficients for technetium in clayey and sandy site soil to be acceptable for the same reason. The effect of changing the technetium sorption coefficients all to a value of approximately zero (i.e., reducing the values by a factor of 1 x 1030) did not change the projected dose detectably from changing only the iodine sorption coefficients. The NRC staff expects that result occurred because the sorption coefficients for technetium in subsurface soil used in the Compliance Case in the 2020 SDF PA are low (Table 2), and other radionuclides did not make a significant contribution to the projected dose to an off-site member of the public in the 2020 SDF PA.
The NRC staff finds the leachate impact factors the DOE applied in the 2020 SDF PA to be acceptable because the sensitivity analysis of changing all the sorption coefficients by a factor of 1 x 1030 bound the potential effects of the leachate impact factors and produced a small (i.e., 10%) increase in the projected peak dose to an offsite member of the public within 10,000 years of site closure. Although the NRC staff determined that uncertainty remains about the
applicability of the experiments conducted with Hanford sediments and simulated leachate-impacted groundwater to leachate-impacted soil at SRS, that uncertainty is not risk-significant in the 2020 SDF PA. The NRC staff finds the extent of the soils to which the DOE applied the leachate impact factors to be acceptable because the bounding alternatives of applying the factors to all soils or no soils produced either a negligible increase or a decrease in the projected peak dose to a member of the public (Figure 11).
4.2.4 Associated Monitoring Factors MF 10.09: Kd Values for SRS Soil As discussed in Section 4.2 of this TRR, the NRC staff determined that the sorption coefficients the DOE used to model I and Tc sorption in site soil were acceptable for the purpose of demonstrating compliance with the performance objectives in the 2020 SDF PA because reasonable alternatives had a minimal effect on the projected dose. However, the DOE has not demonstrated that the chemical conditions of the F-Area samples the DOE used to measure I sorption coefficients are applicable to the SDF. More significantly, the NRC staff determined that the DOE has not demonstrated either: (1) a sound basis for applying leachate impact factors from experiments with Hanford soils to SDF soils, or (2) a sufficient basis for retaining values based on Hanford sediments when experiments conducted with SDF sediments provided significantly different, and more conservative, values. Therefore, although the NRC staff determined that the effects of reasonable changes in the I sorption coefficients in site soil or on the leachate impact factors are not risk-significant in the 2020 SDF PA, the NRC staff cannot conclude that the values are well-supported.
The current 2013 NRC Monitoring Plan for the SDF indicates that the NRC staff expects to close MF 10.09, Kd Values for SRS Soil, if the NRC staff determines that the sorption coefficients for risk-significant radionuclides in the next PA revision (i.e., the 2020 SDF PA) are well supported.
Recommendation HFFM-03 Therefore, the NRC staff recommends closing MF 10.09 and opening a new low-priority monitoring factor entitled Kd Values for SRS Soil in MA 7 under both PO §61.41 and PO §61.42. The NRC staff also recommends monitoring leachate impact factor issues and including them in the new monitoring factor.
MF 7.01: Certain Risk-Significant Kd Values in Site Sand and Clay In technical notes supporting MF 7.01 in Appendix A of the current 2013 NRC Monitoring Plan for the SDF, the NRC staff provided the following summary of issues related to the sorption coefficient for Se:
[The] NRC staff determined that the Kd value of 1000 mL/g assumed in the PA for Se in sand and clay was not adequately supported because that value was representative of Se sorption in a low pH soil. An analysis described by DOE in the RAI response (SRR-CWDA-2011-00044), estimated that the Se Kd values will decrease sharply (i.e., there is less sorption) as the pH increases above pH 6; and will decrease an order of magnitude as the pH approaches 7. In the RAI response, DOE also provided pH readings for site groundwater samples. Of those samples, 42% (30 of 72) had a pH value over 6 and 8.3% (6 of 72) had a pH over 7. Furthermore, the pH of water near the disposal structures may be elevated by the presence of saltstone and the cementitious disposal
structures. Additionally, NRC staff believes that the multiple measurements of a Se Kd of precisely 1,041 mL/g in the original report cited by DOE (WSRC-STI-2006-00037, Table
- 6) appears unusual and is suggestive of an experimental artifact that has not been explained by DOE.
The DOE did not address those issues for Se in the 2020 SDF PA or supporting documents.
The DOE measured leachate impact factors of 0.01 (unitless) in SRS sediment in concrete-equilibrated and cement-equilibrated leachate (DOE document SREL Doc. No. R130004).
That result is consistent with the NRC staff expectation that the Kd for Se would be lower in sediment with a pH greater than the pH of the original measurements (i.e., greater than 6 to 7).
However, in contrast with the NRC staff expectation based on the DOE RAI response in the DOE document SRR-CWDA201100044, the DOE chose to use a leachate impact factor of 1 for Se (i.e., no effect of disposal structure and grout leachate, which the NRC staff expects to have a pH of 10 or greater for thousands of years after site closure).
The DOE based the leachate impact factor of 1 on a previously measured a leachate impact factor of 1.4 for Se in an experiment with Hanford sediments (DOE document PNNL166663).
The DOE based that leachate impact factor on sorption values for Se in unimpacted (5 mL/g) and leachate-impacted (7 mL/g) Hanford soil. The basis the DOE provided for continuing to use a leachate impact factor of 1 despite the values less than 0.01 that the DOE calculated based on measurements with SDF soil was that there were a wide range of measured values.
However, the NRC staff expects that values based on measurements with SDF soil should be weighted more heavily than values based on Hanford soil for SDF analyses.
The NRC staff determined that the sorption coefficients for Se in SDF soils are inadequately supported based on: (1) the unresolved issues summarized in the current 2013 NRC Monitoring Plan for the SDF, and (2) the inconsistency of the leachate impact factor for Se with the experimental data from SRS (DOE document SREL Doc. No. R130004). However, based on the DOE sensitivity analysis in which the DOE reduced sorption coefficients in all site soils by a factor of 1 x 1030, the NRC staff determined that the sorption coefficients for Se in site soil are not risk-significant in the 2020 SDF PA. The current 2013 NRC Monitoring Plan for the SDF states that the NRC could close MF 7.01 if it determined that there was sufficient support for the Kd values for Se in site soil or if NRC determines that Se Kd values for SRS sand and clay do not have the potential to significantly affect the dose to an off-site member of the public.
Recommendation HFFM-04 Therefore, the NRC staff recommends closing MF 7.01 (Certain Risk-Significant Kd Values in Site Sand and Clay).
The NRC staff would continue to monitor the development of support for the Kd values for Se and for the leachate impact factors under the new monitoring factor entitled Kd Values for SRS Soil in MA 7, which the staff recommended opening in recommendation HFFM-03.
4.3 Flow and Transport in the Saturated Zone 4.3.1 Regional Model: The GSA Groundwater Flow Model The NRC staff previously documented its review of the 2016 and 2018 GSA models in the 2016 and 2018 GSA Models TRR; however, the emphasis of the review pertained to SRS Tank Farms, specifically the areas around the FTF and HTF. This document relies on the 2016 and
2018 GSA Models TRR for general conclusions made on technical aspects of the 2016 and 2018 GSA models, but not on conclusions specific to the SDF area.
4.3.1.1 Hydrostratigraphy, Model Domain, and Construction The DOE PORFLOW updated execution plan (DOE document SRNL-STI201600261) stated that the DOE intended to evaluate the incremental impact of using the newly refined hydrostratigraphy model and the DOEs complete set of hydrostratigraphic picks were expected to lead to improved model performance (DOE document SRNL-STI201700008). The NRC staff has concluded that DOEs updated hydrostratigraphic model is acceptable because the latest GSA model (2021 GSA model) modifies the tops of the TCCZ and the UTRA-LAZ, thereby more accurately defining the elevation and thickness of the TCCZ near and downgradient of SDS 4 using the most updated hydrostratigraphic picks (DOE document SRR-CWDA202100065).
4.3.1.2 Boundary Conditions, Recharge, and Calibration Targets In addition to the boundary conditions described in Section 3.3.1 of this TRR, the DOE also applied a combined recharge/drain boundary condition to the top boundary using a new option in PORFLOW version 6.42.3 (DOE document SRNL-STI201700008), thereby alleviating the need to respond to simulated hydraulic head and manually switching between recharge and discharge conditions at the top boundary thereby allowing quicker model convergence (DOE document SRNL-TR201500061). The NRC staff finds the assignment of model boundary conditions to be acceptable because some of the conditions were corrective reassignments (e.g., unintentional error in the boundary condition applied to the base of the 2004 GSA model) and others were supported by technical bases.
Currently, baseflow data is used to help validate the model; however, the NRC staff had previously recommended use of baseflow data as a calibration target (ML19277H550) to avoid issues with non-unique solutions associated with use of well water-level calibration targets alone. Information about baseflow is important to understanding whether the model is well calibrated and whether the water budget of the model has significant errors. The NRC staff had previously recommended (ML19277H550) that an attempt be made to use stream baseflow as a calibration target and to improve the baseflow data. This could be achieved as follows:
streamflow data be collected on GSA streams to develop updated baseflow measurements; a period of record be used that matches the time period over which well water-level measurements were averaged for the purpose of developing calibration targets; and an attempt be made to quantify the input from each side of those streams included in the GSA model (the current assumption is that the groundwater input into the streams is evenly split between the two sides of the stream).
There is uncertainty about the appropriateness of using the current baseflow dataset, and comparisons of simulated versus observed baseflow to the UTR show that the calibrated model significantly under-predicts baseflow to the creek. Until the DOE can implement the above changes, the current use of baseflow data as a validation goal is acceptable.
The DOE provided the results of sensitivity cases involving recharge as discussed in Section 3.3.1 of this TRR by showing that a modest change in the recharge rate in a modestly sized area can modestly change peak concentrations. The NRC staff will monitor the support for the GSA model under a newly created monitoring factor entitled Confidence in GSA Modeling
Results in MA 7 (see Section 4.3.1.6 of this TRR). For example, additional confidence in the GSA model results could be gained by performing additional sensitivity analyses involving variations in recharge rates and the size of the area they are applied by differentiating the local area recharge rates based on topographical elevation.
4.3.1.3 Calibration Targets The DOE attempted to reduce uncertainty in calibration targets and improve the calibration process through the review of water-level data and associated uncertainty, of operational activities that may have influenced water-level measurements, and of the variability in rainfall data. The NRC staff also reviewed the distribution and clustering of well locations used for calibration.
Additional activities were recommended in the 2016 and 2018 GSA Models TRR. Such activities included correlating rainfall rates and water levels at representative wells in the GSA, so that hydrographs for representative wells with long-term data, or average water levels over various averaging periods, could be shown to strengthen the basis for the selected time period and support its use of rainfall data as a surrogate for hydraulic head.
As stated in the 2012 NRC SDF TER, calibration statistics for all of the GSA do not provide helpful information on the goodness-of-fit of the model to long-term conditions that are local to SDF. In general, successful calibration that focuses on the waste disposal facility area of interest provides confidence and this also applies to the SDF. While DOE constructed new wells at Z-Area (see Section 3.3 in this TRR) that can be used as calibration targets, the new wells have only been sampled for a few years, so they do not represent long-term average hydraulic head yet. The NRC staff is aware that the DOE has clearly given preference to calibration targets with a longer period of record.
The DOE Z-Area SDF groundwater monitoring reports present separate figures showing periodic water table elevations for the UTRA-UAZ and potentiometric surfaces for the UTRA-LAZ, and the NRC staff views both aquifer zones as separate hydrogeologic units. As extensively discussed in Sections 3.1 and 4.1 of this TRR, the northern part of the Z-Area, the part of the SDF that will contain the cylindrical disposal structures, has no wells in the UTRA-UAZ that can be used as calibration targets. Current calibration targets in that part of the SDF are the new monitoring wells in the UTRA-LAZ which collect data on the potentiometric surface of the aquifer zone.
While the DOE made significant improvements in the calibration process through re-evaluation of calibration targets, additional support for the approach used to develop the calibration targets is needed. The NRC staff will monitor the support for the GSA model under a newly created monitoring factor entitled Confidence in GSA Modeling Results in MA 7 (see Section 4.3.1.6 of this TRR). For example, additional confidence in the GSA model results could be gained by including additional monitoring wells in the area of interest, specifically wells screened in the UTRA-UAZ located in the northern part of the Z-Area.
4.3.1.4 Model Calibration and Validation For the 2018 GSA model, or the GSA_2018.LW model, update, the DOE-recommended 2016 GSA baseline layer-cake, weighted optimization model was recalibrated, and the layers above the Gordon Confining Unit were optimized for hydraulic conductivity distributions using PEST.
Calibration using optimization algorithms is desirable because it can improve the model fit to
data, reduce the calibration effort, and provide statistical information about model uncertainty (DOE document SRNL-TR201500061). The Gordon Confining Unit itself was assigned a vertical hydraulic conductivity value 7.5 times less than the 2016 GSA model value.
Recalibration of the recommended 2016 GSA model used existing H-Area and Z-Area polygons (DOE document SRNL-STI201700008) and a local F-Area and H-Area seepage-basins polygon (DOE document SRNL-STI201800643) for localized material property calibration.
The recalibrated hydraulic conductivities for the Z-Area TZ in the 2018 GSA model created the desired effect of increasing lateral flow and lessening the vertical flow thereby bringing the model into better alignment with the conceptual flow model demonstrated by the SDS 4 plume (i.e., significant flow and transport in the TZ unit). The DOE modeled the calibrated TZ as being more transmissive, that is, the DOE modeled the horizontal hydraulic conductivity as being identical to that of the global TZ zone and was modeled with a hydraulic conductivity that was higher than the local LAZ horizontal hydraulic conductivity (DOE document SRNL-STI201800643). The new 2021 GSA led to further improvements as seen by comparing Figure 21 below (Figure 14 from the DOE document SRR-CWDA202100065) with the 2018 GSA model reverse particle track (Figure 13 from the DOE document SRR-CWDA202100065).
Updated TCCZ stratigraphy near SDS 4 and revised moisture characteristic curves yielded a 2021 GSA model reverse particle track that originates from the SDS 4 Cell G, as seen in Figure 21 below. The NRC staff agrees that this result shows the GSA model has been improved; however, the 2021 GSA model flow field still appears to be inconsistent with the hydrogeologic conceptual model supported by the SDS 4 plume data. Although there appears to be lateral movement in the unsaturated zone (the modeled mechanism for this lateral movement is unclear to NRC staff), significant groundwater lateral flow on top of the TCCZ downgradient of SDS 4 does not appear to be occurring in the model (see discussion in Section 4.1 of this TRR).
Lack of significant modeled flow and transport in the UTRA-UAZ can also be seen in Figure 17.
In this figure, taken from the November 2021 DOE document SRR-CWDA202100072, particle tracking from all the disposal structures show pathlines in the UTRA-UAZ, TCCZ, and the UTRA-LAZ. As can be seen in that figure, significant lateral groundwater transport is not simulated downgradient of SDS 4. The TRR on Future Scenarios and Conceptual Models (ML23017A088) will discuss plausible conceptual flow and transport models and the 2020 SDF PAs modeling efforts in detail.
Before the smaller grid discretization than the GSA models was applied to the ATM, a velocity field was generated directly from the coarser-scale GSA model. A massconserving linear interpolation scheme was then used to assign velocities to the refined horizontal mesh of the ATM. However, these local velocity fields appear to be different than the velocity fields from the GSA model. A comparison between the streamline traces of the GSA model shown in Figure 18 and the streamline traces of the ATM shown in Figure 17 show slight differences in the direction and in the ten-year travel times. It is not clear to NRC staff how much of the travel time is occurring in the unsaturated zone for the GSA model. A look at the reverse particle track in Figure 21 shows that a considerable lateral distance is traveled in the unsaturated zone before entering the saturated zone. As stated above, it is not clear to NRC staff the process by which this simulated lateral movement in the unsaturated zone is occurring. In any case, where GSA water particles in the unsaturated zone come into contact with the water table is not immediately below the disposal structure, but a considerable distance downgradient. This is in comparison to the contaminant flux for the ATM, which is directly below the disposal structure (i.e., no lateral movement in the unsaturated zone). The NRC staff will monitor the local velocity fields in the ATM and compare them to the velocity fields from the GSA model under a newly created monitoring factor entitled Confidence in GSA Modeling Results in MA 7.
Figure 21: Reverse Particle Track from Well ZBG 2 Using the 2021 GSA Model Flow Field (Figure 14 from SRR-CWDA202100065) 4.3.1.5 Uncertainty As discussed in the 2016 and 2018 GSA Models TRR, there is a large uncertainty associated with the hydraulic conductivity of the UTRA-UAZ given the large resulting multiplier range (DOE document SRNL-STI201800643) indicating that multiple sets of parameters are available to match calibration targets. Although insensitivity of the UTRA-UAZ hydraulic conductivity value to overall calibration is possible, it is not likely, since extra UTRA-UAZ hydraulic conductivity zones were created for the Z-Area and H-Area to improve model calibration in multiple modeling campaigns. Additional calibration local to the waste disposal facility area of interest could improve model performance. Because the UTRA-UAZ, which includes the TZ unit, is an important hydrostratigraphic layer, additional new monitoring groundwater wells could help by providing hydrostratigraphic information and well-water-level data for use as calibration targets.
Other approaches were also discussed in the 2016 and 2018 GSA Models TRR, such as electromagnetic borehole flowmeters, which can measure the vertical distribution of the local vertical horizontal conductivity of a layered aquifer along wellbores at SRS.
4.3.1.6 Associated Monitoring Factors
MF 10.10: Far-Field Model Calibration The NRC Monitoring Plan for the SDF indicates that the NRC expects to close MF 10.10 under PO §61.41 after the DOE updates the PA and NRC determines that the far-field model calibration, particularly in the area near the SDF, is adequate. The NRC staff has concluded that the far-field model calibration of the GSA model, particularly in the area near the SDF, is not adequate. In addition to baseflow data from the local area not being available as a calibration target, the NRC staff finds that the number of other calibration targets in the area of interest to be insufficient. Additional monitoring wells screened in the Z-Area are needed. In particular, monitoring wells are needed that are screened in the UTRA-UAZ and located in the northern part of the Z-Area to provide additional data and therefore confidence in the hydraulic conductivity results. Although NRC staff will monitor DOE efforts to calibrate various versions of the GSA model, it will no longer do so under MF 10.10. The NRC staff recommends that some monitoring factors, including MF 10.10, be incorporated into other monitoring factors for clarity and classification purposes. In this case, the NRC staff recommends incorporating the technical issues from MF 10.10 into a new monitoring factor.
New Monitoring Factor in MA 7: Confidence in GSA Modeling Results The NRC staff identified several technical issues associated with the GSA model results worth monitoring. These issues were discussed in Section 4.3 of this TRR and include concerns about a constant recharge rate applied to the region surrounding the SDF and the technical basis for the assumed modeled depth of the unsaturated zone (obtained from the GSA model and used in the local SDF ATM) being weak.
Recommendation HFFM-05 Therefore, the NRC staff recommends opening a medium-priority monitoring factor entitled Confidence in GSA Modeling Results under MA 7 (Subsurface Transport) under PO §61.41. The NRC expects to close the monitoring factor after the NRC determines that the modeled unsaturated thicknesses below the disposal structures, the local recharge rates, and the far-field model calibration, particularly in the area near the SDF, are adequate, and the NRC staff determines that the differences in the local velocity fields and the velocity fields from the GSA model (possibly due to unsaturated lateral flow in the GSA model) are not significant to dose.
Recommendation HFFM-06 The NRC staff recommends closing MF 10.10 (Far-Field Model Calibration) and incorporating the technical issues from MF 10.10 into a new medium-priority monitoring factor entitled Confidence in GSA Modeling Results in MA 7. These technical issues included the concerns about the lack of data from UTRA-UAZ monitoring wells in the northern Z-Area.
4.3.2 Local Model: The SDF Aquifer Transport Model 4.3.2.1 Model Domain and Grid As discussed in Section 3.3.2 of this TRR, a massconserving linear interpolation scheme was used to assign velocities to the refined 7.6 m x 7.6 m (25 ft x 25 ft) horizontal mesh of the ATM thereby reducing the chance of excessive numerical dispersion (DOE document SRNLSTI
201800012). The NRC staff has concluded that the grid resolution chosen to avoid any significant numerical dispersion for longitudinal dispersivities was adequate.
Because the velocity field for the local PA models is generated with a mass-conserving linear interpolation scheme directly from the regional GSA velocity model (ML19277H550), the local transport models do not necessitate a separate flow model with their own boundary conditions and material property assignments. However, in the 2019 NRC SDF OOV Report (ML19143A084), the NRC staff indicated that calibration local to the waste disposal facilities would be expected to greatly improve model performance in the areas of interest.
4.3.2.2 Dispersion, Recharge, and Contaminant Flux During SRS Tank Farms technical reviews, the NRC staff raised technical issues associated with the degree of dispersion in the DOE models (ML112371715 and ML14094A496). The DOE had previously indicated that additional grid refinement may be necessary to reduce numerical dispersion in cases of very low to no assumed physical dispersion (DOE document SRR-CWDA200900054). For example, if no physical dispersion is assumed, then the peak concentrations associated with a pulse release of a conservative tracer are shown to be a factor of approximately three to four times higher with a grid refined by a factor of two in each dimension (or a factor of 8 times more elements). Updates to the dispersion model since then included the upwinding of diffusivities at cell faces versus harmonic averaging and the use of the full dispersion tensor (DOE document SRR-CWDA201700065), as well as use of the four-parameter stratified model, that is, longitudinal horizontal, longitudinal vertical, transverse horizontal, and transverse vertical dispersivities versus use of a simpler two-parameter model (ML19277H550). The 2016 and 2018 GSA Models TRR states that the use of the four-parameter dispersion model appeared to lead to a significant increase in dose. Since the 2012 NRC SDF TER was issued, the ATM grid has also been refined from 15 m x 15 m (50 ft x 50 ft) to 7.6 m x 7.6 m (25 ft x 25 ft). The NRC staff finds the grid discretization and numerical dispersion for the ATM to be acceptable due to the change in the approach for determining the influence of dispersivity.
Current DOE plans for the SDF include engineered surface covers that are expected to considerably reduce the groundwater recharge below the cover and disposal structures. The NRC staff finds that the DOE analysis to determine the effect of reduced groundwater recharge due to engineered surface covers by simulating reduced groundwater recharge rates provides useful insights (see Section 3.3.2). Results showed that increased dose estimates were relatively small. However, until the closure caps have been constructed and the actual hydraulic conditions have been established, the reduction in groundwater recharge under the cover and the potential effects of that reduction are uncertain, and the NRC staff will continue to review and monitor information on this topic under the newly created MF entitled Confidence in Local SDF Modeling Results under MA 7 (see Section 4.3.2.3 of this TRR).
The colloid transport sensitivity cases assumed that radionuclides would contact and sorb to naturally occurring microscopic clay colloids while remaining suspended in the water resulting in much faster transport than the retarded radionuclides in equilibrium with immobile solids. The NRC staff finds the DOEs evaluation to determine the effect of colloid transport on the total dose to be acceptable because the sensitivity cases showed the influence of colloid facilitated transport on the total dose are negligible due to the lack of highly sorbing radionuclides being transported.
4.3.2.3 Associated Monitoring Factors MF 10.11: Far-Field Model Source Loading Approach The NRC Monitoring Plan for the SDF indicates that the NRC expects to close MF 10.11 under PO §61.41 after DOE updates the PA and NRC determines that the far-field source loading approach in the model is adequate. The NRC staff has concluded that the far-field source loading approach in the ATM is adequate because the local SDF GSA grid has been further discretized from 15 m x 15 m (50 ft x 50 ft) to 7.6 m x 7.6 m (25 ft x 25 ft) since the 2012 NRC SDF TER was issued. Figure 16 in this TRR shows the model domain of the ATM, including the refined grid, and that each disposal structure contains numerous, equally spaced nodes for source loading in the ATM.
Recommendation HFFM-07 The NRC staff recommends that MF 10.11 be closed.
MF 10.12: Far-Field Model Dispersion The NRC Monitoring Plan for the SDF indicates that the NRC expects to close MF 10.12 under PO §61.41 after DOE updates the PA and when the NRC staff determines that the grid refinement used in any hydrological model supporting the updated PA does not increase modeled dispersion beyond the expected physical dispersion. The NRC staff has concluded that the grid discretization and numerical dispersion for the ATM to be acceptable due to the use of the four-parameter dispersion model and the refinement of the grid to a resolution intended to avoid any significant numerical dispersion for longitudinal dispersivities as low as 3.0 m (10 ft)
(SRNLSTI201800012).
Recommendation HFFM-08 The NRC staff recommends that MF 10.12 be closed.
Confidence in Local SDF Modeling Results The NRC staff identified several technical issues associated with the local SDF ATM results worth monitoring. These issues were discussed in Section 4.3.2 and include concerns about differences in these local velocity fields from the ATM appear to be different than the velocity fields from the GSA model possibly due to unsaturated lateral flow in the GSA model and a concern with expected significant reduction in the recharge to the saturated zone due to the closure cap and disposal structures.
Recommendation HFFM-09 Therefore, the NRC staff recommends opening a new medium-priority MF entitled Confidence in Local SDF Modeling Results in MA 7 (Subsurface Transport) under PO
§61.41. The NRC expects to close the new factor when the DOE has constructed the closure caps and future flow conditions underneath the SDF are better known, and when the NRC staff determines what affect those future flow conditions will have on dose.
Impact of Calcareous Zones on Contaminant Flow and Transport The NRC staff did not receive updated information or review new information related to the impact of calcareous zones on contaminant flow and transport so that the low priority status of the monitoring factor related to this topic remains unchanged. Detailed information on the
impact of calcareous zones with respect to site stability is in the NRCs TRR entitled Site Stability (ML23017A114).
Recommendation HFFM-10 The NRC staff recommends closing MF 10.13 and opening a new low-priority monitoring factor entitled Impact of Calcareous Zones on Contaminant Flow and Transport in MA 7 (Subsurface Transport) under the performance objectives of §61.41 and §61.42.
5.0 Teleconference or Meeting There were no teleconferences or meetings with the DOE related to this TRR.
6.0 Follow-up Actions There are no specific Follow-up Actions related to this TRR. The NRC staff will continue to monitor groundwater and far-field modeling under the monitoring factors listed below in Section 7.
7.0 Conclusions The NRC staff concluded that the models used for simulating unsaturated and saturated flow and transport in the 2020 SDF PA are adequate for modeling the projected dose from the SDF for the purpose of the DOE demonstrating compliance with the 10 CFR 61.41 performance objective, Protection of the General Population from Releases of Radioactivity if there is not significant lateral flow and radionuclide transport in the UTRA-UAZ. The NRC staff made that conclusion because these models and their parameter values were generally consistent with the hydrogeological knowledge of the SDF area; however, the models to not represent the potential for significant lateral flow in the UTRA-UAZ. The NRC staff noted other areas in which the model might not be representative in Sections 4.2 and 4.3 of this TRR; however, the staff determined that those areas do not affect the NRC staffs ability to use the model results to determine compliance with 10 CFR 61.41. A conceptual model that incorporates significant lateral flow and radionuclide transport in the UTRA-UAZ, and simulating that conceptual model of flow and transport, will be discussed in the technical review report on Future Scenarios and Conceptual Models (ML23017A088).
The NRC staff concluded that the current groundwater monitoring system for the SDF is not adequate in order to assess whether leaching from saltstone at each disposal structure has occurred. The NRC staff made that conclusion because the number of groundwater monitoring wells in the UTRA-UAZ is not adequate. The NRC staff previously made a similar conclusion in the 2018 TRR entitled, Technical Review of Groundwater Monitoring at and Near the Planned Saltstone Disposal Facility. The most plausible conceptual model of flow and transport suggests that both aquifer zones need to have monitoring wells to adequately monitor potential releases from disposal structures. In addition, NRC staff has determined that the UTRA-Lower Aquifer Zone (LAZ) has insufficient background wells; therefore, the NRC staff may be unable to assess if leaching from saltstone into the UTRA-LAZ is occurring.
The recommendations made by the NRC staff in this TRR are captured in the monitoring factors below. These monitoring factors are listed and briefly described as well as changes made to their status and priority. New-created monitoring factors are also listed and briefly described.
Some monitoring factors listed may have been relabeled for clarity and classification purposes and are described below:
Recommendation HFFM-01 Subsurface Flow and Transport The NRC staff recommends that MA 7 (Subsurface Transport) be relabeled as MA 7 (Subsurface Flow and Transport) because flow is a risk-significant factor that requires monitoring.
Recommendation HFFM-02 Modifying Identification and Monitoring of Groundwater Plumes in the Z-Area The NRC staff recommends modifying the high-priority MF 8.03 (Identification and Monitoring of Groundwater Plumes in the Z-Area) under MA 8 (Environmental Monitoring) under the performance objectives of §61.41 and §61.42 by adding specific issues related to the groundwater monitoring system in the Z-Area as described in this TRR. Consequently, the NRC staff expects to close MF 8.03 when the NRC staff determines that the groundwater monitoring system in the Z-Area can: (1) identify saltstone contaminants in the groundwater in the SDF at no more than 150 ft [46 m] from a disposal structure; (2) identify saltstone contaminants separately in both the UTRA-UAZ and the UTRA-LAZ; (3) obtain background concentration measurements from monitoring wells in the UTRA-UAZ and the UTRA-LAZ upgradient of the SDF; and (4) track the movements of groundwater plumes.
Recommendation HFFM-03 Kd Values for SRS Soil The NRC staff recommends closing the low priority MF 10.09 and opening a new low-priority monitoring factor entitled Kd Values for SRS Soil in MA 7. The NRC staff also recommends monitoring leachate impact factor issues under the new monitoring factor.
Recommendation HFFM-04 Closing Certain Risk-Significant Kd Values in Site Sand and Clay The NRC staff recommends closing MF 7.01, Certain Risk-Significant Kd Values in Site Sand and Clay. The NRC staff will continue to monitor the development of support for the Kd values for selenium and for the leachate impact factors under the new monitoring factor Kd Values for SRS Soil in MA 7, previously MF 10.09 (Kd Values for SRS Soil).
Recommendation HFFM-05 Confidence in GSA Modeling Results The NRC staff recommends opening a medium-priority monitoring factor entitled Confidence in GSA Modeling Results under MA 7 (Subsurface Transport) under the performance objectives of §61.41 and §61.42. The NRC expects to close the new monitoring factor when the NRC staff determines that the modeled unsaturated thicknesses below the disposal structures, the local recharge rates, and the far-field model calibration (previously MF 10.10), particularly in the area near the SDF, are adequate, and the differences in the local velocity fields and the velocity fields from the GSA model are not significant to dose.
Recommendation HFFM-06 Closing Far-Field Model Calibration The NRC staff recommends closing MF 10.10 and incorporating the technical issues from MF 10.10 into a new monitoring factor entitled Confidence in GSA Modeling Results, which the staff recommended opening in a HFFM-05. These technical issues
include the concerns about the lack calibration targets in the UTRA-UAZ as discussed in this TRR.
Recommendation HFFM-07 Closing Far-Field Model Source Loading Approach The NRC staff recommends closing MF 10.11 because the NRC staff has concluded that the modeled far-field source loading approach is adequate.
Recommendation HFFM-08 Closing Far-Field Model Dispersion The NRC staff recommends closing MF 10.12 because the NRC staff has concluded that the grid discretization and numerical dispersion for the ATM is acceptable.
Recommendation HFFM-09 Confidence in Local SDF Modeling Results The NRC staff recommends opening a new medium-priority monitoring factor entitled Confidence in Local SDF Modeling Results in MA 7 (Subsurface Transport) under the performance objectives of §61.41 and §61.42. The NRC expects to close the new monitoring factor when the DOE has constructed the closure caps and future flow conditions underneath the SDF are better known, and when the NRC staff determines what affect those future flow conditions will have on dose.
Recommendation HFFM-10 Modifying Impact of Calcareous Zones on Contaminant Flow and Transport The NRC staff recommends closing MF 10.13 and opening a new low-priority monitoring factor entitled Impact of Calcareous Zones on Contaminant Flow and Transport in MA 7 (Subsurface Transport) under the performance objectives of §61.41 and §61.42.
8.0 References South Carolina Department of Health and Environmental Control (SCDHEC), Regulation 61-107.19 SWM: Solid Waste Landfills and Structural Fill, May 23, 2008. ML101600010.
U.S. Department of Energy (DOE), PNNL16663, Geochemical Processes Data Package for the Vadose Zone in the Single-Shell Tank Waste Management Areas at the Hanford Site, September 2007. ML16106A149
___, SREL Doc. No. R130004, Ver. 1.0, Impact of Cementitious Material Leachate on Contaminant Partitioning, September 2013. ML14196A195
___, SRNL-STI-2012-00518, Radioiodine Geochemistry in the SRS Subsurface Environment, May 2013. ML14196A201.
___, SRNL-STI-2016-00261, General Separations Area (GSA) Groundwater Flow Model Update: Program and Execution Plan, April 2016. ML18107A108.
___, SRNL-STI-2017-00008, Groundwater Flow Simulation of the Savannah River Site General Separations Area, September 2017. ML18081A304.
___, SRNL-STI-2017-00445, Impacts of Updated GSA Groundwater Flow Models on the FTF, HTF and SDF PAs, September 6, 2017. ML18081A308.
___, SRNLSTI201800012, Recommended Aquifer Grid Resolution for EArea PA Revision Transport Simulations, January 2018. ML23058A002
___, SRNL-STI-2018-00336, Updated General Separations Area (GSA) Groundwater Model Calibration Targets, July 2018. ML20206L167
___, SRNLSTI201800643, Updated Groundwater Flow Simulations of the Savannah River Site General Separations Area, January 2019. ML19053A383
___, SRNS-RP-2014-01214, Z-Area Groundwater Tc-99 Characterization Sampling Plan, January 2015. ML15027A371
___, SRNS-RP-2015-00902, Z-Area Groundwater Characterization Data Report, January 2016.
___, SRNS-TR-2014-00283, Z-Area Saltstone Disposal Facility Groundwater Monitoring Report for 2014, January 2015. ML15027A371.
___, SRNS-TR-2015-00300, Z-Area Saltstone Disposal Facility Groundwater Monitoring Report for 2015, January 2016. ML16057A150.
___, SRNS-TR-2016-00110, Z-Area Saltstone Disposal Facility Groundwater Monitoring Midyear Report for 2016, July 2016. ML16365A097.
___, SRNS-TR201700387, Rev. 0, Z-Area Saltstone Disposal Facility Groundwater Monitoring Report for 2017, January 2018. ML18066A363.
___, SRNS-TR-2019-00326, Z-Area Saltstone Disposal Facility Groundwater Monitoring Report for 2019, January 2020. ML20154K593.
___, SRNS-TR-2021-00966, Z-Area Saltstone Disposal Facility Groundwater Monitoring Report for 2021, January 2022. ML22027A001.
___, SRR-CWDA-2014-00099, Rev. 1, Comment Response Matrix for NRC Staff Request for Additional Information on the Fiscal Year 2013 Special Analysis for the Saltstone Disposal Facility at the Savannah River Site, January 2015. ML15020A672.
___, SRR-CWDA201600004, Rev. 1, Comment Response Matrix for NRC Staff Request for Additional Information on the Fiscal Year 2014 Special Analysis for the Saltstone Disposal Facility at the Savannah River Site, March 2016. ML16105A043
___, SRR-CWDA201600052, Rev. 1, Presentation for Savannah River Site Salt Waste Disposal NRC Onsite Observation Visit, April 2016. ML16134A185
___, SRR-CWDA201700068, Evaluation of Impacts to FTF and HTF PA Doses Due to the Update of the GSA Database, October 2017. ML18081A322
___, SRR-CWDA201900001, Performance Assessment for the Saltstone Disposal Facility at the Savannah River Site, March 2020. ML20190A056
___, SRR-CWDA202100065, Response to NRC Request for Supplemental Information #8:
Upper Three Runs Aquifer - Upper Aquifer Zone Lateral Flow Analysis, August 2021.
___, SRR-CWDA202100072, Comment Response Matrix for the Second Set of U.S. Nuclear Regulatory Commission Staff Requests for Additional Information on the Performance Assessment for the Saltstone Disposal Facility at the Savannah River Site, August 2021.
___, WSRCSTI200600198, Hydraulic Property Data Package for the EArea and Z-Area Soils, Cementitious Materials, and Waste Zones, September 2006. ML101600380
___, WSRC-TR950227, Z-Area Saltstone Disposal Facility Groundwater Monitoring Report, June 1995. ML23058A001
___, WSRC-TR9900248, Regional Groundwater Flow Model for C, K, L, and P Reactor Areas, Savannah River Site, Aiken, SC, September 1999. ML19056A205.
___, WSRC-TR200400106, Groundwater Flow Model of the General Separations Area Using PORFLOW, July 2004. ML12185A205.
___, WSRC-TR200500257, Rev. 5, Groundwater Monitoring Plan for the Z-Area Saltstone Disposal Facility, July 2010. ML14196A220
___, WSRC-TR200800001, Z-Area Groundwater Monitoring Report for 2007, January 2008.
U.S. Nuclear Regulatory Commission (NRC), 2012 Technical Evaluation Report for the 2009 Performance Assessment for the Saltstone Disposal Facility at the Savannah River Site, May 2012. ML121170309.
___, U.S. Nuclear Regulatory Commission Plan for Monitoring Disposal Actions Taken by the U.S. Department of Energy at the Savannah River Site Saltstone Facility in Accordance with the National Defense Authorization Act for Fiscal Year 2005, Rev. 1, September 2013.
___, NRC February 4 - 5, 2015, Onsite Observation Visit Report for the Savannah River Site Saltstone Disposal Facility, May 2015. ML15041A562
___, NRC July 7 - 8, 2015, Onsite Observation Visit Report for the Savannah River Site Saltstone Disposal Facility, September 2015. ML15236A299
___, NRC April 19 - 21, 2016, Onsite Observation Visit Report for the Savannah River Site Saltstone Disposal Facility, July 2016. ML16147A197
___, Technical Review of Groundwater Monitoring at and Near the Planned Saltstone Disposal Facility, May 2018. ML18117A494
___, Supplement To The 2013 U.S. Nuclear Regulatory Commission Saltstone Disposal Facility Monitoring Plan, October 2018. ML18219B035
___, U.S. Nuclear Regulatory Commission March 18-19, 2019, Onsite Observation Visit Report for the Savannah River Site Tank Farms, August 2019. ML19143A084
___, U.S. Nuclear Regulatory Commission September 17, 2019, Onsite Observation Visit Report for the Savannah River Site Saltstone Disposal Facility, November 2019. ML19289A525
___, Technical Review of the General Separations Area 2016 and 2018 PORFLOW Models and Associated Documentation Supporting the F-Area and H-Area Tank Farm Facility Performance Assessments at the Savannah River Site, Aiken, SC, December 2019. ML19277H550
___, Technical Review: Inventory for the DOE Performance Assessment for the Saltstone Disposal Facility at the Savannah River Site, March 2022. ML22076A128
___, Technical Review: Future Scenarios and Conceptual Models for the 2020 Performance Assessment for the Saltstone Disposal Facility at the Savannah River Site, Rev. 1, April 18, 2023. ML23017A088
___, Technical Review: Site Stability for the 2020 Performance Assessment for the Saltstone Disposal Facility at the Savannah River Site, Rev. 1, April 18, 2023. ML23017A114