ML21119A317

Enclosure 1 - Technical Review Report Environmental Monitoring Report - PROJ0734

ML21119A317
Person / Time
Site:	PROJ0734
Issue date:	05/14/2021
From:	Cynthia Barr Division of Decommissioning, Uranium Recovery and Waste Programs
To:	Stephen Koenick Division of Decommissioning, Uranium Recovery and Waste Programs
Barr C
Shared Package
ML21119A312	List: ML21119A316 ML21119A317
References
Download: ML21119A317 (29)

Text

TECHNICAL REVIEW OF 2019 ENVIRONMENTAL MONITORING AND NATURAL ATTENUATION REPORTS FOR SAVANNAH RIVER SITE F-AREA AND H-AREA TANK FARM FACILITIES Date: April 2021 Reviewers:

Cynthia S Barr, Senior Risk Analyst, U.S. Nuclear Regulatory Commission David Pickett, Senior Program Manager, Center for Nuclear Waste Regulatory Analyses1 Cynthia Dinwiddie, Principal Scientist, Southwest Research Institute2 Peer Reviewers:

George Alexander, Risk Analyst, U.S. Nuclear Regulatory Commission Mark Fuhrmann, Geochemist, U.S. Nuclear Regulatory Commission Primary Documents Reviewed:

1. ERD-EN-2005-0127, Scoping Summary for the General Separations Area Western Groundwater Operable Unit (U), Savannah River Nuclear Solutions, Savannah River Site: Aiken, SC. November 2020.

2. SRNL-L3230-2019-00005, Rev. 0, Status of SRNL Radiological Field Lysimeter Experiment (RadFLEx) - Year 7, Savannah River National Laboratory: Aiken, SC.

January 2020.

3. SRNS-RP-2020-00064, Savannah River Site, Environmental Report 2019, Savannah River Nuclear Solutions, Savannah River Site: Aiken, SC. 2020.

4. SRNS-RP-2020-00066, Rev. 0, 2019 Annual Groundwater Monitoring Report for the F-and H-Area Radioactive Liquid Waste Tank Farms (U), Savannah River Nuclear Solutions, Savannah River Site: Aiken, SC. March 2020.

5. SRNS-RP-2021-00401, Rev. 0, 2020 Annual Groundwater Monitoring Report for the F-and H-Area Radioactive Liquid Waste Tank Farms (U), Savannah River Nuclear Solutions, Savannah River Site: Aiken, SC. March 2021.3

6. SRRA021685-000012, Rev. A, Partitioning of Cesium-137 and Other Gamma-Emitting Radionuclides to SRS Sediments Recovered from Field Lysimeter Experiments at the Savannah River Site, Prepared by Clemson University, Environmental Engineering and Earth Sciences: Anderson, SC, in partial fulfillment of the scope of work outlined in SRR PO SRRA021685SR. July 2020.

1The Center for Nuclear Waste Regulatory Analyses (CNWRA) is a federally funded research and development center, which was established in 1987 by the U.S. Nuclear Regulatory Commission. The CNWRA is part of the Chemistry and Chemical Engineering Division of Southwest Research Institute, San Antonio, Texas 78238.

2 The Southwest Research Institute is a not-for-profit research institute benefiting government, industry, and the public through innovative science and technology, located in San Antonio, Texas 78238.

3 SRNL-RP-2021-00401, 2020 Annual GW Monitoring Report for the F/H Area Radioactive Liquid Waste Tank Farms (U), was completed just prior to issuance of this technical review report. While the focus of this technical review report is on the 2019 report, any significant changes noted by NRC staff during their cursory review of the 2020 report is also included in this report.

Enclosure 1

7. SRRA021685-000013, Rev. A, Determination of Constituent Concentrations in Field Lysimeter Effluents: FY19 Report, Prepared by Clemson University, Environmental Engineering and Earth Sciences: Anderson, SC, in partial fulfillment of the scope of work outlined in SRR PO SRRA021685SR. August 2020.

8. WSRC-RP-2000-4134, Scoping Summary for the General Separations Area Eastern Groundwater Operable Unit (U), Savannah River Nuclear Solutions, Savannah River Site: Aiken, SC. November 2020 2

Summary and Evaluation:

This technical review report is an update to three previous reports on the same topic dated December 17, 2019, April 20, 2018, and March 31, 2015 (Agencywide Documents Access and Management System [ADAMS] Accession Nos. ML19280A059, ML18051B154, and ML12272A124) with the former reports evaluating the F-Area Tank Farm (FTF) monitoring well network (ML12272A124), H-Area Tank Farm (HTF) monitoring well network (ML18051B154),

source of non-volatile beta/Technetium (Tc)-99 plume at well FTF 28 (ML19280A059), as well as lysimeter and natural attenuation studies conducted for the tank farm facilities (TFFs).

This technical review is associated with Monitoring Factors (MFs) 4.1, Natural Attenuation of Key Radionuclides, and 4.3, Environmental Monitoring, listed in the NRCs combined F- and H- Tank Farm Facility monitoring plan entitled U.S. Nuclear Regulatory Commission Plan for Monitoring Disposal Actions Taken by the U.S. Department of Energy (DOE) at the Savannah River Site (SRS) F-Area and H-Area Tank Farm Facilities in Accordance with the National Defense Authorization Act for Fiscal Year 2005, issued in October 2015 and available using ADAMS Accession No. ML15238A761. MF 4.1 is focused on U.S. Nuclear Regulatory Commission (NRC) staff review of studies that provide information on key radionuclide mobility in the natural system to provide support for key assumptions in DOEs performance assessment (PA) important to DOEs compliance demonstration for the TFFs. MF 4.3 is focused on NRC staff review of environmental monitoring data collected by DOE that also provides information about tank farm releases and provides data to assess DOEs flow and contaminant transport modeling.

This technical review report is divided into two parts. The first part is focused on NRC staff review of environmental monitoring reports prepared by DOE SRS for the FTF and HTF that provide information on radionuclide mobility in the subsurface at SRS. The second part is focused on review of results of lysimeter studies that also provide information on the potential for SRS soils and cementitious materials to retain key radionuclides in the engineered and natural systems at SRS.

Part I: Summary and Evaluation of Environmental Monitoring Results 2019 Monitoring Report for Tank Farms (SRNS-RP-2020-00066)

FTF Monitoring Summary The Environmental Protection Agency (EPA) and the South Carolina Department of Health and Environmental Control (SCDHEC) approved a Sampling and Analysis Plan (SAP) for F-Area Tank Farm (FTF) in December 2012 (SRNS-RP-2012-00287, Rev. 1). During scoping of the monitoring strategy and development of the sampling plans, US DOE, US EPA, and SC DHEC identified gaps in the monitoring well network. DOE subsequently added new wells at both FTF and HTF. Only after closure of the FTF can additional wells be placed due to the presence of active utilities and operating facilities.

During 2019, SRS recorded 1,321 mm (52 in) of precipitation at the H-Area weather station, which was greater than the reported 30-yr average of 1,194 mm (47 in) per year.

Also, 1,118 mm (44 in) of rainfall were recorded in the 3rd and 4th quarters of 2018, alone.

Greater than average water-levels were also noted. Average water-levels at the FTF were reported for the upper aquifer zone (UAZ) and lower aquifer zone (LAZ) of the Upper Three 3

Runs aquifer at values of 67.4 m (221 ft) (UAZ) and 64.0 m (210 ft) (LAZ) in SRNS-RP-2020-00066 for the 2018 reporting period. These water-levels represent 0.3 m (1 ft) above long-term average levels for the UAZ and 0.9 m (3 ft) above average for the LAZ.

The number of FTF wells by aquifer zone are as follows: UAZ (7); LAZ (4) and 2 background wells (one each in the UAZ [FBG-1D] and the LAZ [FBG-1C]). Figure 1 illustrates the locations of the FTF monitoring wells. In the 1st and 3rd quarters of CY2019, 12 of 13 wells were sampled at FTF (FBG-1D was dry as it has been since it was installed 20124). Samples were analyzed for gross alpha, non-volatile beta, tritium [i.e., hydrogen (H)-3], nitrate/nitrite, cadmium (Cd),

chromium (Cr), manganese (Mn), and sodium (Na). Additionally, technetium (Tc)-99 was analyzed due to its known presence in the groundwater. As provided in the SAP, if screening results for gross alpha or non-volatile beta exceed trigger levels of 15 pCi/L (555 Bq/m3) and 50 pCi/L (1,850 Bq/m3), respectively, then contingent analyses for specific radionuclides will be performed by DOE.

The analytical results from monitoring during 2019 and 2020 were similar to previous years.

In 2019, wells FTF-20, FTF-28 and FTF-12R exceeded a screening trigger level (gross alpha screening level of 15 pCi/L (555 Bq/m3) at FTF-28 and non-volatile beta screening level of 50 pCi/L (1,850 Bq/m3) at FTF-20, FTF-28 and FTF-12R). Because the concentrations exceeded the alpha or beta screening levels, contingency analyses were performed by DOE to analyze for specific radionuclides.

Tritium concentrations at well FTF-30D have been variable. Tritium concentration trends (Figure 8 in SRNS-RP-2020-00066 Rev. 0) are replotted in Figure 2 with water-level data.

Concentrations of H-3 in FTF-30D were lower during 2019 sampling. To try and explain the variability in H-3 concentrations, an analysis was performed to determine if changes in precipitation or water-levels at well FTF-30D were responsible for the changes in H-3 concentration. As illustrated in Figure 2 and 3 below, no apparent correlation was noted.

NRC staff plotted data that were included in the 2012 to 2020 environmental monitoring reports to evaluate a potential correlation in water-level with H-3 concentration. An unexpected upward trend in water-level over time was noted at well FTF-30D (Figure 2). To determine the significance of the upward trend, a Seasonal-Kendall test was performed in Visual Sample Plan (VSP). The results indicate an upward trend with a false positive error rate of 5 percent.

The regression line for precipitation also had a very low but positive slope during this time period as illustrated in Figure 3. It is unclear if the increase in water-levels at FTF tanks can be solely attributed to increases in precipitation. To investigate this trend further, nearby wells were also plotted (see Figure 4).

Many of the wells from the 2012 to present timeframe appear to have increasing water-level trends. Additionally, Figure 4 seems to indicate that water-levels declined in years leading up to 2012 and then water-levels at wells sampled after 2012 had a subsequent increase to present day. Many of the historical wells sampled prior to 2012 were located in closer proximity to the heart of the FTF and generally had higher water-levels compared to FTF wells that are currently being sampled. Figure 5 shows the locations of the current FTF wells that are part of the 2012 SAP (primarily located downgradient (west) of FTF), as well as wells that had historically been sampled at FTF (primarily located within the FTF).

4 SRS is evaluating alternative locations for the installation of a new UAZ background well.

4

Figure 1 FTF Monitoring Well Locations. Image Credit: SRNS-RP-2020-00066, Figure 4.

Water-level (primary) and tritium (H-3) concentration (secondary) at Well FTF-30D vs. time water level tritium concentration 224 120 223 100 222 221 80 220 219 60 218 40 217 216 20 215 214 0 11/18/2010 4/1/2012 8/14/2013 12/27/2014 5/10/2016 9/22/2017 2/4/2019 6/18/2020 10/31/2021 Figure 2 Water-level (primary axis) in feet and H-3 concentration (secondary axis) in pCi/mL versus time at Well FTF-30D. There are 3.28 feet/meter (ft/m), and 37 Bq/m3 per 1 pCi/L.

5

Individual Chart for All Locations Observations 1-106 12 11 10 9

UCL = 8.468 8

Analyte 1 7

6 5

Mean = 4.253 4

3 2

1 0 LCL = 0.038 1/1/2012 1/1/2014 1/1/2016 1/1/2018 1/1/2020 Time All Locations; All Seasons 12 Regression Slope = 0.0231688 / year, Rho=0.0251531 11 10 9

8 Analyte 1 7

6 5

4 3

2 1

0 1/1/2012 1/1/2014 1/1/2016 1/1/2018 1/1/2020 Time Regression Line 12 All Locations; All Seasons 11 10 9

8 Analyte 1 7

6 5

4 3

2 1

0 1/1/2012 1/1/2014 1/1/2016 1/1/2018 1/1/2020 Time LOWESS Plot Figure 3 Barnwell monthly precipitation data in inches/month (in/mo) and regression curves.

Note: The top figure shows a control chart (two standard deviations) from the Parameter-elevation Regressions on Independent Slopes Model (PRISM) data set. The middle figure shows a linear regression curve, while the bottom figure shows a Locally Weighted Scatterplot Smoothing (LOWESS) fit to the precipitation data. Created using Visual Sample Plan software.

6

Figure 4 Historical and more recent water-levels at FTF wells.

Note: Obvious outliers on the high end were filtered from the data set to allow for clearer presentation of the data.

The top figure shows historic wells that were sampled from late 1980s to present. The bottom figure shows more recent sampling, including wells sampled since the SAP was approved in 2012. Created using Tableau software.

7

Water-levels are an important input (i.e., often serve as calibration targets) to the saturated groundwater flow model used to simulate contaminant fate and transport at the TFFs and water-level variability could have a significant impact on tank farm performance. NRC will follow-up with DOE regarding the potential cause(s) of the apparent water-level decreases and increases at FTF and how they impact calibration targets and contaminant fate and transport modeling.

Figure 5 FTF current and historical well locations.

8

SRNS-RP-2020-00066 also notes that the non-volatile beta results seemed to be reversed relative to normal conditions at FTF-20 [108 pCi/L (3,996 Bq/m3)] versus FTF-12R [13.7 pCi/L (507 Bq/m3)] (see Figure 6). DOE reasoned the results had been reversed because FTF 12R had been higher since 2014 but was lower and representative of FTF-20 data in 2019.

Data collected in 2020 and reported in SRNS-RP-2021-00401 also supports the FTF-20 and FTF-12R data had been reversed.

Consistent with previous reporting, FTF-28 had the highest concentrations of non-volatile beta

[590 and 746 pCi/L (21,830 and 27,602 Bq/m3)] in the 2019 report. Technetium-99 is the primary source of the non-volatile beta plume. Tc-99 has previously been detected in wells FTF-28 and FTF-12R and was at a maximum of 1640 pCi/L (60,680 Bq/m3) at FTF-28 in 2019.

The concentration of Tc-99 can be higher than non-volatile beta contribution from Tc-99 due to loss of activity from volatilization in the drying step of the non-volatile beta analysis.

DOE continued to attribute the source of Tc-99 at FTF-28 to releases from the F-Area Inactive Process Sewer Line (FIPSL). Elevated hydrogen ion concentration [or relatively low pH (negative log of hydrogen ion concentration)] at FTF-28 (pH 5.07) is noted by DOE as evidence that Tc-99 is sourced by FIPSL. DOE also notes other potential sources are past releases and contamination areas within the FTF facility boundary (SRNS, 2012a). NRC staff had previously provided results of backwards particle tracking conducted by DOE at the request of NRC, which showed that the source of contamination from LAZ well, FTF-28, was likely further upgradient (i.e., sourced from within the FTF footprint) and therefore, not likely to be sourced by the FIPSL located downgradient from FTF and in close proximity to FTF-28 (ADAMS Accession No. ML19280A059). Iodine (I)-129 was also detected at FTF-28, FTF-19 and FTF-12R.

NRC staff will continue to monitor the source of the FTF plumes as it is important to understanding contaminant flow and transport processes at the FTF.

9

Tc-99/Non-volatile Beta I-129 Figure 6 FTF wells with non-volatile beta/Tc-99 and I-129 contamination.

10

Table 1 Summary of FTF Monitoring Well Results as of 2019*

Nitrate- H-3 Gross Na Non- I-129 Tc-99 nitrite Alpha volatile beta Max 7.07 3.14 28.5 19,000 746 pCi/L 1.5 pCi/L 1640 pCi/L Concentration mg/L pCi/mL pCi/L§ µg/L (2019)

Wells FBG-1C FTF-30D FTF-28 FTF-19 FTF-28 FTF-28 FTF-28 (max well LAZ FTF-19 FTF-30D FTF-20 FTF-12R FTF-12R FTF-12R top) (bkgd) FTF-20 (UAZ) FTF-19 FTF-22 FTF-20 (LAZ)

FSL-11C (LAZ)

No. of 29 of 29 All wells 18 of 30 28 of 28 22 of 30 ND 9 of 17 samples samples samples samples samples samples detected Historic high 7.4 105 pCi/mL 217 pCi/L 33,300 959 pCi/L 1.91 pCi/L 1670 pCi/L mg/L µg/L Historic high FTF-19 FTF-30D FTF-18 FTF-30D FTF-28 FTF-28 FTF-28 well (UAZ) (UAZ) (1/21/99, (UAZ) (LAZ) (LAZ) (LAZ)

(2003 to 2015 (2014) (2/21/2017) per 2009 (2013) (2012) (2013) (2017) unless report) otherwise noted)

Source Potentially Potentially Potentially 1961 FIPSL or FIPSL or release 1961 1961 near release release Tank 8 near near (SRNS-RP- Tank 8 Tank 8 2012-00287, Rev. 1)

Note: 1 pCi/L = 37 Bq/m3.

• The monitoring report presenting 2020 data (SRNS-RP-2021-00401) was published just prior to issuance of this technical review report and is therefore, not the focus of this report. However, a cursory review of the 2020 data is provided.

Nitrate-nitrite was detected a little higher at 8.4 mg/L in background well FBG-1C (LAZ) in 2020 (SRNS-RP-2021-0041).

Tritium was detected at 6.3 pCi/mL (233,100 Bq/m3) in FTF-12R in 2020 (SRNS-RP-2021-00401).

§ The maximum concentration that was not estimated (j qualified) was 4.3 pCi/L (159 Bq/m3) at FTF-30D.

Qualified/estimated value.

Scoping Summary for the General Separations Area Western Groundwater Operable Unit ERD-EN-2005-0127 In 2019, 33 monitoring wells, 4 seepline piezometers, and 3 surface water sampling stations were monitored as part of the General Separations Area Western Groundwater Operable Unit.

Groundwater constituent concentrations remained consistent with results from 2018 and groundwater plumes were relatively stable. The report also noted that surface waters were not impacted above EPA drinking water standards (i.e., maximum contaminant limits). Continued sampling of the western groundwater operable unit monitoring network was recommended.

Three distinct areas are the focus of the General Separations Area Western Groundwater Operable Unit monitoring well network, including the north and west plumes. The focus of NRCs review is on the south plume, which may be sourced by the FIPSL or sources within the FTF facility boundary. The south plume consists of H-3, non-volatile beta, I-129, Strontium 11

(Sr)-90, Radium (Ra)-226, and Tc-99. Non-volatile beta exceeded the screening level of 50 pCi/L (1,850 Bq/m3) in 5 of 13 wells. The maximum concentration of non-volatile beta, 474 pCi/L (17,538 Bq/m3), was detected in well FTF-28, which is also sampled as part of the FTF monitoring well network. Maximum concentrations for a number of constituents are provided in the table at the bottom of Figure 8.

Figure 7 depicts the non-volatile beta plume that extends from FTF toward Upper Three Runs in the south plume. The absence of contamination at well FSL-6D (see Figure 7) may suggest two different sources of non-volatile beta in the subsurface downgradient from FTF.

Nearby well FSL-7D does have non-volatile beta contamination and is located in close proximity to a section of the FIPSL that is known to have collapsed. These results may suggest that the portion of the plume downgradient of FSL-7D is sourced by the FIPSL, while Tc-99 in well FTF-28 and other downgradient wells may be sources by FTF. This hypothesis is also supported by particle tracking conducted by DOE at the request of NRC. Backwards particle tracking from FTF-28 resulted in particles terminating near Type I tanks (i.e., near Tank 8) at FTF (ADAMS Accession No. ML19280A059).

Figure 7 South non-volatile beta plume in GSA Western Groundwater Operable Unit downgradient from FTF. Image credit: Figure 7 in ERD-EN-2005-0127 (2020).

Figure 8 depicts the location of key radionuclides of concern in the Western Groundwater Operable Unit along with maximum concentrations. Strontium-90 is only present in one south plume well, FSL-7D, perhaps due to its limited mobility and association with the FIPSL collapse.

It is also interesting to note that gross alpha is detected in wells FSL-7D and FGS-12C.

Higher mobility H-3, nitrate, and Tc-99 seem to be more pervasive in the south plume, perhaps due to enhanced dispersion compared to other less mobile radionuclides. Analysis of groundwater monitoring data is helpful to better understanding key radionuclide mobility and contaminant flow and transport processes at the FTF. NRC staff will continue to evaluate groundwater monitoring data generated by DOE for the TFFs.

12

Figure 8 Constituents present in south plume in Western Groundwater Operable Unit [Adapted from Figure 2 in ERD-EN-2005-0127 (2020)]. There are 37 Bq/m3 per 1 pCi/L.

13

HTF Monitoring Well Summary Similar to FTF, the SAP for HTF was approved by the EPA and SCDHEC in December 2012 (SRNS-RP-2012-00146, Rev. 1). During scoping of the monitoring strategy and development of the sampling plans, US DOE, US EPA, and SC DHEC identified gaps in the monitoring well network, so DOE subsequently added new wells at both FTF and HTF. Due to active utilities and operating facilities, new wells cannot be added to the HTF monitoring well network until after closure of the HTF (and FTF).

Above (30-year) average precipitation rates recorded at the H-Area weather station in 2019 are discussed in the preceding FTF section. Greater than average water-levels were also reported for the UAZ and LAZ of the Upper Three Runs aquifer at values of 82 m (269 ft) (UAZ) and 77 m (252 ft) (LAZ). These water-levels represent 0.6 m (2 ft) above long-term average levels for the UAZ and LAZ at HTF.

The number of HTF wells by aquifer zone are as follows: UAZ (17); LAZ (28), Gordon Aquifer Unit GAU (1) and 3 background wells (one each in the UAZ [FBG-1D] and the LAZ [FBG-1C]).

See Figure 9 for the locations of the HTF monitoring wells. In the 1st and 3rd quarters of CY2019, all 46 wells were sampled at HTF. Samples were analyzed for gross alpha, non-volatile beta, Tc-99, H-3, nitrate/nitrite, Cd, Cr, Mn, and Na. As provided in the SAP, if screening results for gross alpha or non-volatile beta exceed trigger levels of 15 pCi/L (555 Bq/m3) and 50 pCi/L (1,850 Bq/m3), respectively, then contingent analyses for specific radionuclides will be performed by DOE. In 2019, no results exceeded the screening levels for gross alpha or non-volatile beta.5 DOE reported that the monitoring results were similar to previous years. Results indicated low concentrations of nitrate-nitrite and H-3 in most wells. A summary of the monitoring well results for 2019 as well as historical data is presented in Table 2.

5 SRNS-RP-2021-00401 reports that, in 2020, gross alpha at HAA-12B and nonvolatile beta at HAA-7C exceeded the trigger levels.

14

Figure 9 HTF monitoring well network.

15

Table 2 Summary of HTF Monitoring Well Results as of 2019*

Nitrate- H-3 Gross Alpha Na Non-volatile Tc-99 nitrite beta Max 6.39 mg/L 47.1 pCi/mL 7.99 pCi/L 19,200 µg/L 21.1 pCi/L§ 12 pCi/L¶ Concentration Most less (2019) than 1 mg/L Wells HAA-4D HAA-12C HAA-18D HAA-11B HAA-4D HAA-15C (max well Similar to HAA-12D (LAZ) top) previous HAA-8D years (UAZ)

No. of 111 out of 70 out of 108 37 of 108 101 out of 59 of 108 1 of 109 samples 113 samples 101 samples samples detected Historic high 34.8 mg/L 355 pCi/mL 29.1 pCi/L 128,000 µg/L 277 pCi/L 88.2 pCi/L Historic high HAA-4D HTF-12 HAA-4D HAA-9AR HAA-9AR HAA-1D well (2003 to (UAZ) (2018) (1986) (UAZ) (GA) (GA) (2010) (UAZ) 2015 unless (9/1/2009) (8/23/2010) Thought to be (3/12/2014) otherwise HAA-12 more anomalous/

noted) recently suspect exhibits according to relatively high 2011 report values Source Off-site Fuels Potentially Potentially Receiving associated associated Basin with 1960 soil with 1960 soil process release near release near sewer lines Tank 16 Tank 16.

(HIPSL)

Note: 1 pCi/L = 37 Bq/m3

• The monitoring report presenting 2020 data (SRNS-RP-2021-00401) was published just prior to issuance of this technical review report and is therefore, not the focus of this report. However, a cursory review of the 2020 data is provided.

In 2020, the maximum gross alpha measurement was 26 pCi/L (962 Bq/m3) at HAA-12B.

In 2020, the maximum Na measurement was 27,700 µg/L in well HAA-12B, which also had the highest gross alpha measurement.

§In 2020, the maximum concentration of non-volatile beta was 55 pCi/L (2,035 Bq/m3) in well HAA-7C.

Upon resampling, the non-volatile beta concentration was only 5 pCi/L (185 Bq/m3) and thought to be associated with naturally occurring materials.

Although the non-volatile beta value of 277 pCi/L (10, 249 Bq/m3) was thought to be anomalous, HAA-9AR also had a relatively high reading of 73.7 pCi/L (2,727 Bq/m3) in 2009 (SRNS-TR-2010-00012), although the value was an unqualified result that did not undergo post-laboratory verification/validation. In 2016, a measured value of 223 pCi/L (8,251 Bq/m3) for HAA-8D was reported, but this was also considered anomalous (SRNS-RP-2017-00073). In 2014, a value of 54.7 pCi/L (2,024 Bq/m3) was reported at HAA-20C (SRNS-RP-2015-00069).

¶In 2020, the maximum concentration of Tc-99 was 14 pCi/L (518 Bq/m3) in well HAA-12C.

Scoping Summary for the General Separations Area Eastern Groundwater Operable Unit WSRC-RP-2000-4134 Tritium is detected in most HTF wells. Figure 10 shows the extent of the H-3 plume, which is thought to be sourced by the H-Area Off-site Fuels Receiving Basin. Tritium is also present in deeper LAZ wells, as depicted in Figure 4 of WSRC-RP-200004134 (2000). NRC staff will continue to monitor the source of the H-3 plume at HTF.

16

Figure 10 Tritium plume at HTF. Image credit: WSRC-RP-2000-4134 (2000), Figure 2.

2019 SRS Annual Report (SRNS-RP-2020-00064)

In 2019, the estimated potential radiation exposure to an offsite representative person at the Savannah River Site boundary (several miles from FTF and HTF) was 0.0018 mSv (0.18 mrem). The 2019 value was lower than the 2018 value of 0.0027 mSv (0.27 mrem).

Of the potential 0.0018 mSv (0.18 mrem) dose, 0.0016 mSv (0.16 mrem) was attributed to liquid releases, and 0.0018 mSv (0.018 mrem) was attributed to air releases. Site discharges to air and water were reported to be below regulatory standards for the public and the environment.

The 2019 decrease in estimated dose was due to (i) reduced H-3 releases from the SRS and (ii) increased Savannah River annual flow volume. The estimated offsite dose is based on measured concentrations or is calculated based on release/transport data and considers all sources of radioactivity at the assessment point. Therefore, the doses bound the potential dose to an offsite receptor from TFF activities, alone.

17

Other doses reported were to the onsite hunter and fisherman. The onsite hunter is exposed from consumption of deer and wild hog meat from annual hunts conducted at SRS.

The estimated dose from harvested deer and hog meat is 0.175 mSv (17.5 mrem). This dose is for an actual hunter who harvested four animals (three hogs and one deer) during the 2019 hunt. For the hunter-dose calculation, SRS conservatively assumes that this hunter individually consumed the entire edible portion of these animals, about 91 kg (200 lbs) of meat.

The maximum potential dose from fish consumption (i.e., bass) at Lower Three Runs was estimated at 0.00227 mSv (0.227 mrem). The hypothetical fisherman is assumed to consume 24 kg (53 lbs) of bass caught exclusively from the mouth of Lower Three Runs.

Liquid Releases Tritium accounts for more than 99 percent of the total amount of radioactivity released from SRS to the Savannah River. In 2019, a total of 452 Ci (1.67 x 1013 Bq) of H-3 were released to the river. SRS measures concentrations of H-3 in the river water and cesium (Cs)-137 in fish at several locations along the Savannah River to estimate dose to off-site members of the public.

Due to the high bioaccumulation factor of Cs-137 in fish, the fish consumption pathway is an important dose pathway. The fish pathway accounts for 63 percent of the estimated potential dose to the off-site representative person from liquid releases. About 25 percent of the potential dose is from consumption of vegetables assumed to have been grown on-site and meat and milk from animals raised on Savannah River water from near the mouth of Steel Creek.

The drinking water pathway accounts for 5 percent of the potential dose. The primary radionuclides contributing to liquid pathway dose are Cs-137 (69 percent) and Tc-99 (11 percent).

Air Releases Tritium accounts for most of the estimated potential dose from air release pathways (i.e.,

67 percent). Iodine-129 accounts for 23 percent of the dose. The primary exposure pathways from air releases include vegetable consumption (39 percent), inhalation (30 percent), and cow milk consumption (22 percent).

To determine the potential dose, calculated radionuclide concentrations are used instead of measured concentrations due to uncertainty in measuring air concentrations at the site perimeter and off-site locations. Although calculated radionuclide concentrations are used to estimate potential doses, measured concentrations are used to validate the dose calculations.

The doses estimated from calculated air concentrations are 1.5x to 2.5x more conservative than values determined using measured H-3 concentrations.

Conclusion of Environmental Monitoring Reviews Previous staff conclusions remain valid and include the following:

1. DOE has performed environmental monitoring that provides useful information on the hydrogeological systems at FTF and HTF. This information can also be used to better understand contaminant flow and transport at the TFFs and provide support for DOE PA models.

2. Uncertainty about the source(s) of contaminant plumes detected via the FTF and HTF monitoring well networks exists. A better understanding of contaminant flow and 18

transport processes at the TFFs, and more extensive data analysis and interpretation could help reduce this uncertainty.

3. PA modeling and analysis should be better integrated with the groundwater monitoring program at the TFFs. For example, FTF and HTF monitoring well placement could be better optimized to detect releases from the TFFs if releases occur in the future.

PORFLOW groundwater contaminant fate and transport models of the TFFs are available but are not being used to design the monitoring well network, particularly to inform vertical placement of wells. PA modeling assumptions and results could be used to determine key constituents and field monitoring data, which would provide the most useful information for evaluating performance of and to detect early releases from the TFFs.

NRC findings with respect to environmental monitoring report reviews include the following:

1. DOE should provide stronger support for the assumed sources of contaminant plumes to ensure that it is able to detect releases from the TFFs.

2. DOE should leverage groundwater monitoring data to obtain information about natural attenuation of key radionuclides at the TFFs.

3. DOE should analyze groundwater monitoring data in greater detail to increase understanding of processes important to contaminant flow and transport and model calibration (e.g., analyze water-level data to better understand water-level response to changes in precipitation, to develop calibration targets for PA models, and to better understand tank farm performance).

19

Part II: Summary and Evaluation of Lysimeter Studies This technical review is associated with MF 4.1, Natural Attenuation of Key Radionuclides, listed in the NRCs combined F- and H- Tank Farm Facility monitoring plan (ADAMS Accession No. ML15238A761). MF 4.1 is focused on NRC staff review of studies that provide information on key radionuclide mobility in the natural system to provide support for key assumptions in DOEs PA important to DOEs compliance demonstration for the TFFs. Three 2020 reports on DOE-sponsored lysimeter experiments at SRS were reviewed. The tests are designed to provide site-specific, in-situ information about factors affecting radionuclide transport in SRS soils. The three reports document three different types of activities: (i) facility refurbishment and lysimeter installation, (ii) leach tests on soil specimens from dismantled lysimeters, and (iii) radionuclide concentrations in lysimeter effluents.

SRNL-L3230-2019-00005, Revision 0 - Status of SRNL Radiological Field Lysimeter Experiment (RadFLEx) - Year 7 This report contains no experimental results; rather, it details how in 2019 the RadFLEx facility was refurbished (treated for rust abatement), and how 11 new lysimeter sources were developed and installed. The three radionuclide source types were:

- Neptunium (Np)-237 as solid Np(IV)O2 or sorbed on goethite, collected on glass fiber filters

- Ra-226 dried from solution onto a glass fiber filter, and

- Plutonium (Pu) (chiefly Pu-239 by mass) as solid Pu(VI)O2CO3, dried on glass fiber filter.

Preparation of the Ra-226 sources was straightforward; it can be assumed that Ra dried from nitric acid will precipitate as Ra(NO3)2. For the Np and Pu sources, the investigators carefully prepared and characterized the solid sources, such that there is reasonable confidence in their starting compositions. For instance, the NpO2 was analyzed by scanning electron microscope (SEM), Scanning Transmission Electron Microscopy (STEM), and selected area electron diffraction to conclusively confirm the solid composition. Appropriate radioanalysis was used to quantify the radionuclide content of each source filter.

The lysimeter studies should provide useful information on radionuclide mobility in the subsurface. The utility of the experiments is increased when the sources and materials used are representative of the real system. For example, NRC staff recognizes that the chemical form and oxidation state of key radionuclides can control radionuclide mobility in the subsurface.

In particular, results of high-level waste release experiments show the potential for higher solubility Pu species to be released from Tank 18F. NRC staff would support DOE efforts to study natural attenuation of higher solubility forms of Pu released from the tanks in these lysimeter experiments. However, DOE did not provide a clear basis for the radionuclide source solids selections. Information about radionuclide source solids selections would have been helpful to include in this report to allow NRC staff to evaluate the relevance of the source selections to tank farm closure. NRC findings from review of SRNL-L3230-2019-00005, Rev. 0 include the following:

The report did not explain the rationale for installing new Np, Ra, and Pu sources, nor for the particular choices of radionuclide source solids. This type of information would have been helpful for providing context for the work that was performed.

20

It will be important, when interpreting lysimeter data in the future, to evaluate how relevant the starting source solids are to understanding radionuclide release and transport from waste forms and grouted tanks at SRS.

The Pu solid identification requires clarification. In the main report text, X-Ray Diffraction (XRD) results were shown to identify the solid as NH4PuO2CO3*x H2O, whereas Appendix B indicates that the final product is PuO2CO3. Furthermore, the report did not explain why the Pu sources were not subjected to the same extensive solids characterization as were the Np sources. Source-phase identification could be important to interpreting Pu lysimeter release behavior.

SRRA021685-000012, Revision A - Partitioning of Cesium-137 and Other Gamma-Emitting Radionuclides to SRS Sediments Recovered from Field Lysimeter Experiments at the Savannah River Site This Clemson University study performed radionuclide desorption experiments on soil samples taken from both an older experiment (M2) begun in 1981 that studied the release from a fragment of vitrified HLW emplaced in a field-scale lysimeter buried in the vadose zone and from a more recent RadFLEx facility (L26) lysimeter, both at A-Area at the SRS, with L26 serving as a control. Soil samples were taken from the recovered lysimeters from 20 to 28 cm (7.9 to 11.0 in) below the radionuclide source for M2 and from 4 to 7 cm (1.6 to 2.8 in) below the source for L26. Radioactivity that had leached from the sources and sorbed onto vadose sediments below the sources were used for these desorption experiments. The radionuclides studied include cobalt (Co)-60, cesium (Cs)-137, Barium (Ba)-133, and Europium (Eu)-152.

The starting experimental solution was 0.01 M NaCl and measured pH after reaction with the soil samples was approximately 4.9. The measured dissolved radionuclide concentration was used with the concentration in untreated sediment to calculate Kd. The studied fission products were measured in leachates and in dried, unleached sediment samples using gamma spectroscopy with a high-purity germanium detector. Measurements were made using consistent geometry and were calibrated using an Eu-152 standard. The studys radioanalytical methods were appropriately applied.

Of the studied radionuclides, only Cs-137 was detected in a M2 sediment sample leach solution, yielding a Kd of 2,050 mL/g. This was in the range of values for the L26 lysimeter (i.e., 1,900 to 2,570 mL/g) but much higher than measured in short-term laboratory studies. For example, the report indicates that the Cs-137 Kd values for the L26 lysimeter are 1000x larger than those reported from seven-day laboratory sorption experiments with similar SRS sediments.

The researchers suggest that increased contact time (i.e., 5 years versus 7 days) results in stronger sorption of Cs-137 to mineral surfaces and that the aging process has been linked to the slow exchange and fixation of Cs-137 onto interlayer (sorption) sites in clay minerals.

NRC staff also note that the Cs-137 and Eu-152 remaining on the soils in this study could be reflective of association with less accessible or less reversible sorption sites. NRC staff also note that a potential hysteresis effect should be considered. The appropriate models and model parameters to use in PAs for long-term disposal of radioactive waste to capture the complexity of radionuclide transport based on the results of these studies should be carefully thought out and well supported by the data.

The L26 lysimeter source had been prepared by drying a solution of these four radionuclides onto a glass fiber filter. Co-60 Kds ranged from 19 to 39 mL/g. These relatively low values are 21

consistent with observations of Co-60 in lysimeter effluent and with short-term laboratory studies. L26 Ba-133 Kds were similarly low at 25 to 31 mL/g, but these values were about 10x higher than expected. Eu-152 Kds from L26 were similar to or higher than Cs-137 values, ranging from 2,180 to 7,590 mL/g (with a large error on the largest value), and were 4x to 10x higher than in laboratory studies.

Field-based sorption experiments can provide useful information on the environmental behavior of radionuclides at SRS. The experiments should be well designed to maximize the usefulness of the results. The results suggest the relative mobility of Co and Ba, and the relative immobility of Cs and Eu. While these radionuclides are not expected to be risk-drivers for SRS Tank Farm closure, as in the case of laboratory Kds, the values obtained from this study should not be applied indiscriminately in PA models. Consideration must be given to the appropriateness of the experimental solutions and conditions, as well as to the lysimeter conditions.

Specific comments on the experiments, their results, and their usefulness include the following:

The older M2 lysimeter leached radioactivity from a glass waste form and the lysimeter experiments begun in 1981 were exposed for 11 years and then capped with a cement barrier for 13 years before being disassembled. The applicability of results from the lysimeters with the vitrified waste form to radioactivity leached from grouted tanks is unclear. Additionally, the conditions of the archived samples that have been stored since 2005 is also unclear and could have affected sorption/desorption behavior of the key radionuclides studied.

Consideration should be given to whether desorption Kds truly reflect a reversible surface process of the sort that is modeled by a partition coefficient. It should be considered whether high Kds could result from a potentially kinetically controlled process, such as precipitation. The report suggests that there may be an aging effect resulting in higher Cs-137 and Eu-152 Kds after longer times, but this hypothesis needs further testing.

The large discrepancy between laboratory and lysimeter Kd values for Cs-137 needs careful study if the results are to be used in PAs. For example, the impact of a possible hysteresis effect should be studied in greater detail prior to use in PAs.

The investigators should consider whether the extraction solutions are chemically representative of percolating waters in the SRS subsurface. While pH values may be appropriate, other chemical constituents can influence sorption behavior.

The investigators should consider using different solid-to-solution ratios in leach tests (the same sediment-to-solution ratio was used; the results of the 1/10 were not reported because they were below detection limits).

SRRA021685-000013, Revision A - Determination of constituent concentrations in field lysimeter effluents This Clemson University study reports radionuclide concentrations in RadFLEx lysimeter effluents collected quarterly over a seven-year period. The emplaced lysimeter radionuclide sources were either (i) prepared oxides, carbonates, or nitrates; (ii) precipitates from dried solutions; or (iii) cementitious materials incorporating radionuclides. Studied radionuclides were Co-60, Cs-137, Ba-133, Eu-152, Np-237, and Pu-239/240/241. These studies provide useful information on the environmental mobility of radionuclides at the SRS.

22

The gamma-emitting radionuclides were measured in the effluents by gamma spectroscopy using a high-purity germanium detector and appropriate methods, similar to those used in SRRA021685-000012 except that larger solution samples were used. The minimum detectable concentrations (MDCs) were appropriately calculated from the lower limit of detection and the characteristics of the gamma energy used for the measurement. Np-237 and Pu-239/240 were measured by inductively coupled plasma mass spectrometry (ICP-MS), using appropriate calibration standards and an internal Pu-242 standard to correct for instrument variation.

MDCs were calculated from ICP-MS calibration curves. For selected samples that had insufficient Pu for detection by ICP-MS, a much more sensitive method was used, involving concentration of Pu by co-precipitation and extraction by ion exchange, followed by alpha counting. The method used a Pu-242 tracer to account for chemical yield and the MDC was appropriately calculated using the background counts and counting efficiencies.

Effluent chemistry, as reflected by pH, was highly variable (pH 3.3 to 8.3) for reasons not well understood, but the overall mean of 5.0 was reasonable for SRS subsurface waters.

Variation of pH within a lysimeter was more limited. The investigators should consider any evidence for pH variability from major element effluent chemistry, which was not discussed in the report. In addition, the dissolved oxygen data should be used with caution because isolating the samples from air was difficult. Another potential source of variability is related to the degree of interaction of infiltrating waters with the soils and cementitious sources. The pathway and residence time of the infiltrating water could vary significantly through the lysimeter.

The distribution of radioactivity after it leaves the source can nonetheless provide useful information regarding radionuclide mobility through the soil column if interpreted carefully.

Of the studied fission products, only Co-60 was detected consistently in effluents. Ba-133 was not detected, and Cs-137 and Eu-152 were detected at low levels during only one sampling event late in the study period. On average, only 0.04 percent of source Co-60 was collected in the effluent. Cementitious sources appeared to consistently have higher effluent Co-60 concentrations than other sources, but the investigators were not sure that was a chemical effect. Notably, effluent from lysimeters with cementitious sources did not consistently have higher pH than effluent from other lysimeters which could be related to mineralogical changes of the cementitious waste forms and associated impacts on permeability. Future studies should try to better elucidate the relationship between effluent chemistry and radionuclide mobility.

Given reported sorption behavior, the investigators did not understand why Co was significantly more mobile than Cs and Ba. NRC agrees that post-test column characterization, including inspection of cementitious source materials, may improve understanding of the transport behavior of these fission products.

The effluent measurements confirmed the expected relative immobility of Pu for at least a significant fraction of the source. In fact, a particularly sensitive analytical method involving Pu co-precipitation and alpha spectroscopy was necessary in order to detect Pu isotopes after ICP-MS was unable to do so. Pu was observed in effluent from only 5 of the 18 lysimeters with Pu in their sources, with measured concentrations ranging from 9x1013 to 1x1015 M.

These low values are reported to be near the solubility limits of Pu(IV) hydroxide phases. If, in fact, the dissolved Pu measured in lysimeter effluent is representative of solubility-controlled release from the source, then transport of the Pu front from the source to the lysimeter bottom in just a few years provides evidence for a possible high-mobility fraction of dissolved Pu.

For example, if one were to assume that the Pu released from the source was solubility-23

controlled [the measured concentrations were stated to be consistent with solubility limits for Pu(IV)], measurement of Pu in effluent after 8 years would suggest a retardation factor of approximately 27 for an assumed infiltration rate of 0.3 m/yr, and moisture content of 0.3.

This would correspond to a Kd of approximately 5 L/kg, which is similar to results from a previous lysimeter study reported in Demirkanli et al. (2007) (see also Kaplan et al., 2006).

This is inconsistent with the very high desorption Kds reported in a December 2020 presentation by Mangold (SRR-CWDA-2020-00088). Discrepancies in retardation estimates based on two methods [i.e., (i) timing of transport through the lysimeter versus (ii) values reported by Mangold (SRR-CWDA-2020-00088)] based on desorption experiments6 requires careful analysis.

An important finding in the report was that decreasing concentrations of Pu with depth were found in the reduced Pu lysimeters as deep as 40 cm (15.7 in) below the surface, suggesting cyclic oxidation and reduction. This conceptual model is consistent with the wetting and drying cycles within the lysimeter soil profile, where penetrating water will cause Pu oxidation and subsequent downward movement. Drying will then cause Pu reduction to the immobile Pu(IV) state. These types of complexities are not currently captured in PA models. DOEs conceptual and numerical models for Pu transport is based on a single, high Kd value. Based on these observations, NRC staff is concerned that DOEs conceptual and numerical models for Pu transport in soils do not adequately represent the limited observations of Pu transport in soils.

DOE should evaluate if the conceptual and numerical models in the PAs adequately represent the distribution of Pu in the lysimeter data. If DOEs conceptual and numerical models are not consistent with those data, DOE should revise their PA conceptual and numerical models accordingly. Based on the risk significance of Pu mobility, DOE may consider additional field data to further reduce uncertainty.

With regard to sample collection, analysis of samples in three-month intervals would lead to averaging of concentrations over the time period of sample collection. Although this might be acceptable for estimation of transport parameters for use in a long-term PA, variability in effluent concentrations that would provide mechanistic information to construct and/or validate conceptual models may be missed. Regardless, it is important that the experimental design preserve information necessary to perform a mass balance on activity (i.e., the remaining activity in the source, the activity in the soils, any activity sorbed to the lysimeter tubes, and the total activity released to the sample collection containers is accounted for) and that the distribution of radioactivity in the spent lysimeters at the end of the experiment is used to better understand radionuclide transport and provide support for PA models.

In contrast to Pu, Np-237 was significantly mobile, with up to 37 percent of the source inventory detected in the effluent. This contrast is consistent with knowledge of the sorption behavior of these two elements. Interestingly, the source Np oxidation state exerts a significant influence on transport behavior. Those lysimeters with Np sources prepared as Np(IV) exhibited much lower effluent Np concentrations than those with Np(V) sources.

6 The Pu desorption experiments were previously reviewed by NRC staff in ADAMS Accession No. ML19280A059.

NRC staff found that the calculated Kds based on desorption experiments did not adequately account for the presence of a more mobile fraction of Pu that could be risk significant. A modeling analysis to determine the Kd value or values that would be required to represent the observed distribution of Pu in the core samples reported in SRRA021685-000009 and SRRA021685-000010 and the effluents reported in SRRA021685-000011 is important to understand if there is a small mobile fraction of plutonium and to reduce uncertainty in the natural attenuation of Pu.

24

NRC agrees that the effluent studies are practical and should continue. The results from this study provide useful information on the relative mobility of radionuclides, but also raise many questions. Key uncertainties in the applicability of these results, with respect to modeling the long-term performance of waste sites at SRS, are:

The causes of the high variability among lysimeters of (i) effluent pH, (ii) concentrations of some radionuclides (e.g., Np), and (iii) quantity of effluent collected The effects of episodic wetting and drying of the soil column The effects of different sources on release rates, effluent chemistry, and radioelement speciation, and The inconsistencies with expected relative fission product behavior based on laboratory studies.

It is understood that studies conducted in natural settings are affected by additional interactions and varying conditions relative to simpler laboratory studies; indeed, that is part of the utility of field studies. The complex interactions of different factors, however, may prevent simple analysis. Nevertheless, understanding should be improved by continued effluent studies and post-test column characterization and leaching studies, but also by more consideration of effluent-water chemistry. DOE should also consider undertaking lysimeter flow and transport modeling. While modeling would not be straightforward, particularly for an unsaturated hydrologic system, modeling may help elucidate the key factors affecting radionuclide migration through the soil column. Ideally, Kd values from lysimeters and lab tests could be validated by determining Kds from field sampling of soil and groundwater in contact with each other, from along plume pathways.

NRC findings with respect to lysimeter report reviews include the following:

1. Although there are limitations to the lysimeter studies (see enumerated list below), the studies provide useful information regarding natural system performance and should be continued to provide support for key modeling assumptions in DOEs PAs.

2. DOE should clarify its source selections and their importance and relevance to tank farm closure.

3. DOE should clarify the chemical form of Pu reported in SRNL-L3230-2019-00005.

4. DOE should clarify the applicability of the SRRA021685-000012 experiments for tank farm closure (e.g., waste form, differences in environmental conditions during storage, solid/solution ratios, solution chemistry, and potential hysteresis effect).

5. The representativeness of the lysimeter study results presented in SRRA021685-000013, Rev. A, to tank farm closure should be discussed (e.g., chemical forms of key radionuclides, environmental conditions), as well as any data gaps pertinent to tank farm performance.

6. The variability in environmental conditions (e.g., precipitation rates through the system) and results should be analyzed and discussed in more detail.

7. DOE should consider modeling the transport behavior of key radionuclides in the natural system using data from the lysimeter studies to provide support for PA models.

8. The applicability of a single equilibrium Kd model to tank farm closure should be critically evaluated. For example, Pu is known to exist in four different oxidation states with substantial variability in mobility depending on its chemical form. The fact that Pu was measured in effluent, albeit at low concentrations, a few years into the lysimeter study suggests that a simple Kd modeling approach does not adequately represent field 25

conditions (i.e., much lower, single digit, Kds could be calculated based on travel time for at least a fraction of the source). As stated in MF 4.1 and Appendix E of the FTF and HTF Monitoring Plan (ADAMS Accession No. ML15238A761), DOE should evaluate the modeling approach used to simulate Pu transport in the subsurface, including use of a single, average Kd. Lysimeter studies provide evidence that a more complex model is necessary to simulate the transport of Pu in the SRS subsurface. To date, NRC staff is unaware of any DOE effort to update the modeling treatment of Pu transport in the natural system to address high-priority MF 4.1 from NRC staffs Monitoring Plan.

This information is needed, however, to assess whether DOE can meet the performance objectives in 10 CFR Part 61, Subpart C for tank farm closure.

Teleconference or Meeting:

No teleconference or meeting was held with DOE related to this TRR.

Follow-up Actions:

NRC staff will follow-up with DOE to better understand factors influencing water-levels at the FTF and HTF that are important to development of calibration targets for the saturated zone flow model, understanding the impact of historical operations on groundwater quality at the TFFs and how that might affect interpretation of groundwater monitoring data, and better understanding the impact of variability on contaminant flow and transport.

NRC staff will also follow-up with DOE regarding DOEs interpretation of lysimeter data and its potential use in updated TFF PAs.

Open Issues:

There are no open issues resulting from this TRR.

==

Conclusions:==

In conclusion, NRC staff will continue to monitor DOE studies related to natural attenuation of key radionuclides and analyze environmental monitoring data to provide support for key PA modeling assumptions. Previous staff conclusions remain valid and include the following:

1. DOE has performed environmental monitoring that provides useful information on the hydrogeological systems at FTF and HTF. This information can also be used to better understand contaminant flow and transport at the TFFs and provide support for DOE Performance Assessment (PA) models.

2. Uncertainty about the source(s) of contaminant plumes detected via the FTF and HTF monitoring well networks exists. A better understanding of contaminant flow and transport processes at the TFFs, and more extensive data analysis and interpretation could help reduce this uncertainty.

3. PA modeling and analysis should be better integrated with the groundwater monitoring program at the TFFs. For example, FTF and HTF monitoring well placement could be better optimized to detect releases from the TFFs should releases occur in the future.

PORFLOW groundwater transport models of the TFFs are available but are not being used to design the monitoring well network, particularly to inform vertical placement of wells. PA modeling assumptions and results could be used to determine key 26

constituents and field monitoring data, which would provide the most useful information for evaluating performance of and to detect early releases from the TFFs.

NRC findings with respect to environmental monitoring report reviews include the following:

1. DOE should provide stronger support for the assumed sources of contaminant plumes to ensure that it is able to detect releases from the TFFs.

2. DOE should leverage monitoring data to obtain information about natural attenuation of key radionuclides at the TFFs.

3. DOE should analyze monitoring data in greater detail to increase understanding of processes important to contaminant flow and transport and model calibration (e.g.,

analyze water-level data to better understand water-level response to changes in precipitation, to develop calibration targets for PA models, and to better understand tank farm performance).

NRC findings with respect to lysimeter report reviews include the following:

1. DOE should clarify its source selections and their importance and relevance to tank farm closure.

2. DOE should clarify the chemical form of Pu reported in SRNL-L3230-2019-00005.

3. DOE should clarify the applicability of the SRRA021685-000012 experiments for tank farm closure (e.g., waste form, differences in environmental conditions during storage, solid/solution ratios, solution chemistry, and potential hysteresis effect).

4. The representativeness of the lysimeter study results presented in SRRA021685-000013, Rev. A, to tank farm closure should be discussed (e.g., chemical forms of key radionuclides, environmental conditions), as well as any data gaps pertinent to tank farm performance.

5. The variability in environmental conditions (e.g., precipitation rates through the system) and results should be analyzed and discussed in more detail.

6. DOE should consider modeling the transport behavior of key radionuclides in the natural system using data from the lysimeter studies to provide support for performance assessment models.

7. The applicability of a single equilibrium Kd model to tank farm closure should be critically evaluated. For example, Pu is known to exist in four different oxidation states with substantial variability in mobility depending on its chemical form. The fact that Pu was measured in effluent, albeit at low concentrations, a few years into the lysimeter study suggests that a simple Kd modeling approach does not adequately represent field conditions (i.e., much lower [single digit] Kds could be calculated based on travel time for at least a fraction of the source). As stated in MF 4.1 and Appendix E of the FTF and HTF Monitoring Plan (ADAMS Accession No. ML15238A761), DOE should evaluate the modeling approach used to simulate Pu transport in the subsurface, including use of a single, average Kd. Lysimeter studies provide evidence that a more complex model is necessary to simulate the transport of Pu in the SRS subsurface. To date, NRC staff is unaware of any DOE effort to update the modeling treatment of Pu transport in the natural system to address high-priority MF 4.1 from NRC staffs Monitoring Plan.

This information is needed, however, to assess whether DOE can meet the performance objectives in 10 CFR Part 61, Subpart C for tank farm closure.

27

References Demirkanli, D.I., F.J. Molz, D.I. Kaplan, R.A. Fjeld, and S.M. Serkiz. Modeling Long-Term Plutonium Transport in the Savannah River Site Vadose Zone. Vadose Zone J. Vol.6, No. 2.

pp. 344-353. 2007. <<https://doi.org/10.2136/vzj2006.0042>>

ERD-EN-2005-0127. Scoping Summary for the General Separations Area Western Groundwater Operable Unit (U). Savannah River Nuclear Solutions. Aiken, South Carolina, November 2020. [ADAMS Accession No. ML19318F627]

Kaplan, D.I., I. Deniz, D.I. Demirkanli, L. Gumapas, B.A. Powell, R.A. Fjeld, F.J. Molz, and S.M. Serkiz. Eleven Year Field Study of Pu Migration from Pu III, IV, and VI Sources.

Environmental Science and Technology. Vol. 40, No. 2. pp. 443-448. 2006.

ML112371751. Technical Evaluation Report for Draft Waste Determination for Savannah River Site F Area Tank Farm. Rockville, Maryland: U.S. Nuclear Regulatory Commission.

October 2011. [ADAMS Accession No. ML112371751]

ML12272A124. Barr, C.S. Technical Review of Environmental Monitoring and Site-Specific Distribution Coefficient Reports. Memorandum to Gregory F. Suber. Rockville, Maryland:

U.S. Nuclear Regulatory Commission. March 2015. [ADAMS Accession No. ML12272A124]

https://www.nrc.gov/docs/ML1227/ML12272A124.pdf ML14094A496. Technical Evaluation Report for Draft Waste Determination for Savannah River Site H Area Tank Farm. Rockville, Maryland: U.S. Nuclear Regulatory Commission.

June 2014. [ADAMS Accession No. ML14094A496]

ML15238A761. U.S. Nuclear Regulatory Commission Plan for Monitoring Disposal Actions Taken by the U.S. Department of Energy at the Savannah River Site (SRS) F-Area and H-Area Tank Farm Facilities in Accordance with the National Defense Authorization Act for Fiscal Year 2005. Rockville, Maryland: U.S. Nuclear Regulatory Commission. October 2015. [ADAMS Accession No. ML15238A761]

ML18051B154. Barr, C.S. Technical Review of Environmental Monitoring Reports for F-Area and H-Area Tank Farm Facilities (PROJ0734). Memorandum to Maria Arribas-Colon.

Rockville, Maryland: U.S. Nuclear Regulatory Commission. April 2018. [ADAMS Accession Nos. ML18051B009, ML18051B153, ML18051B154]

https://www.nrc.gov/docs/ML1805/ML18051B154.html ML19280A059. Barr, C.S. Technical Review of Environmental Monitoring Reports for F-Area and H-Area Tank Farm Facilities (Docket No. PROJ0734). Memorandum to Stephen S.

Koenick. Rockville, Maryland: U.S. Nuclear Regulatory Commission. December 2019.

(ADAMS Accession No. ML19280A059) https://www.nrc.gov/docs/ML1928/ML19280A059.pdf Peruski, K.M., M. Maloubier, D.I. Kaplan, P.M. Almond, and B.A. Powell. Mobility of Aqueous and Colloidal Neptunium Species in Field Lysimeter Experiments. Environmental Science and Technology. Vol. 52. pp. 1963-1970. 2018.

SRNL-L3230-2019-00005. Roberts et al. Status of SRNL Radiological Field Lysimeter Experiment (RadFLEx) - Year 7. Rev. 0. Savannah River National Laboratory:

Aiken, South Carolina. January 2020. [ADAMS Accession No. ML20303A340]

28

SRNS-RP-2020-00064. Savannah River Site, Environmental Report 2019. Savannah River Nuclear Solutions, Savannah River Site: Aiken, South Carolina. 2020. [ADAMS Accession No. ML20303A341]

SRNS-RP-2020-00066. 2019 Annual Groundwater Monitoring Report for the F- and H-Area Radioactive Liquid Waste Tank Farms (U). Rev. 0. Savannah River Nuclear Solutions, Savannah River Site: Aiken, South Carolina. March 2020. [ADAMS Accession No. ML21119A153]

SRNS-RP-2021-00401. 2020 Annual Groundwater Monitoring Report for the F- and H-Area Radioactive Liquid Waste Tank Farms (U). Rev. 0. Savannah River Nuclear Solutions.

Aiken, South Carolina. March 2021. [ADAMS Accession No. ML21085A476]

SRRA021685-000008. Peruski et al. Determination of Constituent Concentrations in Field Lysimeter Effluents. Clemson University under contract with Savannah River Remediation.

Aiken, South Carolina. December 2017. (ADAMS Accession No. ML18067A486)

SRRA021685-000009. Peruski et al. Analysis of Plutonium Soil Concentrations in Field Lysimeter Experiments. Clemson University under contract with Savannah River Remediation.

Aiken, South Carolina. December 2017. (ADAMS Accession No. ML18067A509)

SRRA021685-000010. Peruski et al. Analysis of Plutonium Soil Concentrations in Field Lysimeter Experiments: Soil Pu Concentration Profile from a NH4Pu(V)O2CO3(s) Source.

Clemson University under contract with Savannah River Remediation. Aiken, South Carolina.

October 2018. (ADAMS Accession No. ML19179A062)

SRRA021685-000011. Peruski et al. Determination of Constituent Concentrations in Field Lysimeter Effluents. Clemson University under contract with Savannah River Remediation.

Aiken, South Carolina. October 2018. (ADAMS Accession No. ML19179A047).

SRRA021685-000012. Coutelot et al. Partitioning of Cesium-137 and Other Gamma-Emitting Radionuclides to SRS Sediments Recovered from Field Lysimeter Experiments at the Savannah River Site. Revision A. Clemson University under contract with Savannah River Remediation. Aiken, South Carolina. April 2020. [ADAMS Accession No. ML20303A342]

SRRA021685-000013. Parker et al. Determination of Constituent Concentrations in Field Lysimeter Effluents (FY19 Report). Clemson University under contract with Savannah River Remediation. Aiken, South Carolina. August 2020. [ADAMS Accession No. ML20303A343]

WSRC-RP-2000-4134. Scoping Summary for the General Separations Area Eastern Groundwater Operable Unit (U). Savannah River Nuclear Solutions. Aiken, South Carolina, November 2020. [ADAMS Accession No. ML20350B586]

29