                      						DOCKET: EPA-HQ-OAR-2014-0609
                                       
                                       
                                       
                                       
                                       
                                          
                      Review of EPA Sensitivity Studies of the 
                                  DOE CRA-2014 WIPP
                  Compliance Recertification Performance Assessment
                                          

                                          
                         Contract Number: EP-D-10-042 
                           Work Assignment No. 5-12 
                                       
                 Prepared for:U.S. Environmental Protection Agency 
      Office of Radiation and Indoor Air
Center for Waste Management and Regulations 1301 Constitution Avenue
                                Washington, DC 20004
                      Kathleen Economy: Work Assignment Manager
      Prepared by:S. Cohen & Associates1608 Spring Hill Road Suite 400Vienna, Virginia 22182
                                          
                                      July 2017
PREFACE
The U.S. Department of Energy (the DOE) is required to submit a Compliance Recertification Application (CRA) to the U.S. Environmental Protection Agency (EPA) for the Waste Isolation Pilot Plant (WIPP) facility every five years including an updated assessment of future WIPP performance. During the EPA's review of the DOE's CRA-2014 performance assessment (PA), events associated with the February 2014 repository fire and radionuclide release have resulted in closed portions of the underground facility. This closure has created a situation where certain parts of the underground facility could not be accessed for ground control. Panel 9 may be abandoned along with plans to install panel closures in panels 3, 4, 5 and 6. 
Because the CRA performance assessments are predictions of post-closure repository performance and the EPA knows there will be modifications to the current repository design, modifying the CRA-2014 PA at this time to incorporate alternative parameter values would not add more reality to predictions of repository post-closure performance. Consequently, the EPA adopted the CRA-2014 PA as originally submitted by the DOE as the baseline, rather than have the DOE conduct a revised PA baseline calculation (PABC). In lieu of requesting a PABC-2014, the EPA requested that the DOE the DOE conduct a set of sensitivity studies to address some of the significant technical concerns arising from the EPA's CRA-2014 review. The inputs to these sensitivity studies broadly address many of the EPA's technical concerns that could potentially impact long-term repository performance. The Agency has reviewed the results of these studies and determined that there exists an adequate level of confidence -- that is, a reasonable expectation -- that the repository will continue to comply with EPA regulations. Those results are described and evaluated in this report. 
Additionally, the EPA recommends further work that can be conducted to evaluate many of the technical concerns identified in the EPA's review of the CRA-2014 PA, as well as incorporate future repository design changes. The EPA will work with the DOE to determine the best path forward for resolution of EPA's concerns, which could include additional data reviews, independent technical reviews, and possibly additional sensitivity analyses to reach a consensus for the next CRA. It is anticipated that the results of these efforts will be incorporated into the CRA-2019 PA or otherwise be made available during the EPA's review of the CRA-2019 PA. 

EXECUTIVE SUMMARY
THIS REPORT DOCUMENTS THE RESULTS OF FOUR sensitivity studies conducted by the U.S. Department of Energy (the DOE) at the request of the U.S. Environmental Protection Agency (EPA). The purpose of these studies was to determine the sensitivity of Waste Isolation Pilot Plant (WIPP) CRA-2014 performance calculations to changes in the modeling of fluid flow and geochemical processes within the WIPP underground facility. The studies analyzed the effects of the following model changes.

 Include capillary effects and salt creep closure and healing when simulating fluid flow in the parts of the WIPP repository that do not contain waste,
 Revise the probability of encountering a Castile brine reservoir beneath the repository,
 Revise the uncertainty distributions for actinide solubility,
 Revise the iron sulfidation reactions,
 Use a corrected shear strength distribution for degraded waste,
 Use the corrected version of the DRSPALL code, and 
 Use a corrected panel closure length. 

The sensitivity studies yielded the following principal results and conclusions.
 
BRAGFLO Code Capabilities. The DOE's BRAGFLO fluid flow code is not capable of modeling capillary effects under conditions of changing permeability, and therefore cannot fully simulate the salt creep closure process. This capability would improve WIPP performance modeling and should be incorporated in WIPP performance assessment (PA) as soon as feasible.

Effects of Panel Closure Design Changes. the DOE's decision to replace the concrete monolith panel closure design with a design incorporating run-of-mine (ROM) salt aggregate plugs will, upon salt creep closure and healing, result in much tighter, lower permeability closures that will effectively isolate the waste panels from one-another.

Effects of Creep Closure and Healing. Capillary effects along with low permeability from creep closure and healing of WIPP panel closures and non-waste area drifts effectively eliminates fluid flow among waste panels and results in slightly greater calculated releases than under the open drift assumptions of the CRA-2014 PA. As a result of the changed panel closure design, the open drift assumptions now may underestimate rather than overestimate releases and should be modified for the CRA-2019 WIPP PA. 

Effect on Castile Brine and DBR Modeling. Castile brine reservoir intrusions dominate the occurrence of direct brine releases (DBRs), which control lower probability but higher consequence total repository releases. PA modeling assumptions regarding Castile brine encounters and DBRs should be reevaluated for the CRA-2019 PA in light of changed repository conditions due to the DOE's panel closure redesign.

Effect on Mean Total Releases. Revising the solubility uncertainty distributions, the distribution of GLOBAL:PBRINE, and the iron sulfidation reaction stoichiometric coefficients significantly increased DBRs as well as mean total low probability releases. Using a corrected shear strength distribution for degraded waste, the corrected version of the DRSPALL code, and a corrected panel closure length had negligible effects on total mean releases but should nevertheless be incorporated in the CRA-2019 PA. EPA anticipates that the DOE's update for the CRA-2019 PA will address the issues identified in EPA's review of the CRA-2014 PA.  

Effect on Regulatory Compliance. The mean total releases did not exceed EPA's WIPP release limits, nor the 95% confidence level in that mean, or by any individual vectors in any of the sensitivity studies.

TABLE OF CONTENTS

PREFACE	ii
EXECUTIVE SUMMARY	iii
1.0 INTRODUCTION	1
2.0 SENSITIVITY STUDY OVERVIEW	2
3.0 SEN1 REVIEW	2
3.1 SEN1 Implementation	2
3.2 SEN1 Results	4
3.3 EPA SEN1 Conclusions	5
4.0 SEN2 REVIEW	5
4.1 SEN2 Implementation	5
4.2 SEN2 Results	6
4.3 EPA SEN2 Conclusions	9
5.0 SEN3 REVIEW	9
5.1 SEN3 Implementation	9
5.2 SEN3 Results	11
5.3 EPA SEN3 Conclusions	17
6.0 SEN4 REVIEW	18
6.1 SEN4 Overview	18
6.2 SEN4 Implementation	21
6.3 SEN4 Results	23
6.4 EPA SEN4 Conclusions	24
7.0 COMPLIANCE POINT REVIEW	25
8.0 CONCLUSIONS	26
8.1 Conclusions from SEN1	26
8.2 Conclusions from SEN2	27
8.3 Conclusions from SEN3	27
8.4 Conclusions from SEN4	29
8.5 Summary of Conclusions	30





                                LIST OF FIGURES

Figure 1. SEN2 and CRA-2014 waste panel pressure (WAS_PRES) for a borehole intrusion at 350 years that does not intersect a Castile brine reservoir.
Figure 2. SEN2 and CRA-2014 cumulative brine flow (BRNWASIC) into the waste panel for a borehole intrusion at 350 years that does not intersect a Castile brine reservoir.
Figure 3. SEN2 and CRA-2014 waste panel pressure (WAS_PRES) for a borehole intrusion at 350 years that also intersects a Castile brine reservoir.
Figure 4. SEN2 and CRA-2014 cumulative brine flow (BRNWASIC) into the waste panel for a borehole intrusion at 350 years that also intersects a Castile brine reservoir.
Figure 5. SEN2 and CRA-2014 overall mean CCDFs for direct brine releases.
Figure 6. SEN2 and CRA-2014 overall mean CCDFs for spallings releases.
Figure 7. Comparison of overall means for SEN2 release components.
Figure 8. SEN2 and CRA-2014 overall mean CCDFs for total normalized releases.
Figure 9. Individual total normalized release vectors for the three replicates of the SEN2 study.
Figure 10. SEN2, SEN3 and CRA-2014 waste panel pressure (WAS_PRES) for the undisturbed repository.
Figure 11. SEN2, SEN3 and CRA-2014 cumulative brine flow (BRNWASIC) into the waste panel for the undisturbed repository.
Figure 12. SEN2, SEN3 and CRA-2014 brine saturation (WAS_SATB) in the waste panel for the undisturbed repository.
Figure 13. SEN2, SEN3 and CRA-2014 pressure in the waste panel (WAS_PRES) for a borehole intrusion at 350 years that does not intersect a Castile brine reservoir.
Figure 14. SEN2, SEN3 and CRA-2014 cumulative brine flow (BRNWASIC) into the waste panel for a borehole intrusion at 350 years that does not intersect a Castile brine reservoir.
Figure 15. SEN2, SEN3 and CRA-2014 brine saturation (WAS_SATB) in the waste panel for a borehole intrusion at 350 years that does not intersect a Castile brine reservoir.
Figure 16. SEN2, SEN3 and CRA-2014 pressure in the waste panel (WAS_PRES) for a borehole intrusion at 350 years that also intersects a Castile brine reservoir.
Figure 17. SEN2, SEN3 and CRA-2014 cumulative brine flow (BRNWASIC) into the waste panel for a borehole intrusion at 350 years that also intersects a Castile brine reservoir.
Figure 18. SEN2, SEN3 and CRA-2014 brine saturation (WAS_SATB) in the waste panel for a borehole intrusion at 350 years that also intersects a Castile brine reservoir.
Figure 19. SEN2, SEN3 and CRA-2014 pressure in the waste panel (WAS_PRES) for the S6-BF scenario of a double borehole intrusion at 1,000 and 2,000 years.
Figure 20. SEN2, SEN3 and CRA-2014 cumulative brine flow (BRNWASIC) into the waste panel for the S6-BF scenario of a double borehole intrusion at 1,000 and 2,000 years.
Figure 21. SEN2, SEN3 and CRA-2014 brine saturation (WAS_SATB) in the waste panel for the S6-BF scenario of a double borehole intrusion at 1,000 and 2,000 years.
Figure 22. SEN2, SEN3 and CRA-2014 overall mean CCDFs for direct brine releases.
Figure 23. Comparative average pressure and brine saturations in the modeled waste panel for four BRAGFLO modeling scenarios.
Figure 24. SEN2, SEN3 and CRA-2014 overall mean CCDFs for spallings releases.
Figure 25. Comparison of overall means for SEN3 release components.
Figure 26. SEN2, SEN3 and CRA-2014 overall mean CCDFs for total normalized releases.
Figure 27. Individual total normalized release vectors for the three replicates of the SEN3 study.
Figure 28. Distributions of actinide solubility uncertainty factor (SOLMOD3:SOLVAR) sampled values for the +III oxidation state from the three replicates of the SEN4 and CRA14 studies (from Zeitler and Day 2016, Figure 2-1).
Figure 29. Distributions of actinide solubility uncertainty factor (SOLMOD4:SOLVAR) sampled values for the +IV oxidation state from the three replicates of the SEN4 and CRA14 studies (from Zeitler and Day 2016, Figure 2-2).
Figure 30. Distributions of sampled values of the probability of encountering pressurized Castile brine (GLOBAL:PBRINE) from the three replicates of the SEN4 and CRA14 studies (from Zeitler and Day 2016, Figure 2-3).
Figure 31. Distributions of waste shear strength (BOREHOLE:TAUFAIL) sampled values from the three replicates of the SEN4 and CRA14 studies (from Zeitler and Day 2016, Figure 2-4).
Figure 32. Comparison of spallings and total mean releases using corrected DRSPALL Version 1.22 with earlier, uncorrected versions of DRSPALL.
Figure 33. Overall mean cuttings and cavings releases and confidence intervals for SEN4 and CRA-2014.
Figure 34. Overall mean spallings releases and confidence intervals for SEN4 and CRA-2014.
Figure 35. Overall mean releases to the Culebra and associated confidence intervals for SEN4 and CRA-2014.
Figure 36. Overall mean direct brine releases and associated confidence intervals for SEN4 and CRA-2014.
Figure 37. Comparison of overall means for principal release pathways in the SEN4 analysis.
Figure 38. Total mean releases and associated confidence intervals for SEN4 and CRA-2014.
Figure 39. Total releases for all individual realizations from the three SEN4 replicate analyses.
Figure 40. Mean total releases as a percentage of EPA's upper (0.1 probability) release limit.
Figure 41. Mean total releases as a percentage of EPA's lower (0.001 probability) release limit.

                                LIST OF TABLES

Table 1. the DOE CRA-2014 and EPA Proposed SEN1 Fluid Flow Parameter Values for Operations and Experimental Area Drifts 
Table 2. the DOE CRA-2014 and SEN1 Fluid Flow Parameter Values for DRZ Adjoining Operations and Experimental Area Drifts 
Table 3. the DOE CRA-2014, SEN2 and SEN3 Fluid Flow Parameter Values for Operations and Experimental Area Drifts 
Table 4. the DOE CRA-2014, SEN2 and SEN3 Fluid Flow Parameter Values for DRZ Adjoining Operations and Experimental Area Drifts 
Table 5. the DOE CRA-2014 and EPA Proposed SEN3 Fluid Flow Parameter Values for Waste Panel Closures 
Table 6. the DOE CRA-2014 and EPA Proposed SEN3 Fluid Flow Parameter Values for DRZ Adjoining Waste Panel Closures
Table 7. the DOE CRA-2014 and Actual SEN3 Fluid Flow Parameter Values for Waste Panel Closures 
Table 8. the DOE CRA-2014 and Actual SEN3 Fluid Flow Parameter Values for DRZs Adjoining Waste Panel Closures
Table 9. Comparison of Mean Total Normalized Releases with EPA Release Limit


                             LIST OF ABREVIATIONS


CCA		Compliance Certification Application 
CCDF		Complementary Cumulative Distribution Function
CFR		Code of Federal Regulations
CPR	Cellulosic, Plastic, and Rubber
CRA		Compliance Recertification Application
DBR	Direct Brine Release
the DOE	U.S. Department of Energy
DRZ		Disturbed Rock Zone
EPA		Environmental Protection Agency 
ERMS	Electronic Record Management System
FR		Federal Register
LANL	Los Alamos National Laboratory
LWA	Land Withdrawal Act
OAR		Office of Air and Radiation 
PA 		Performance Assessment 
PABC	Performance Assessment Baseline Calculation
PAVT	Performance Assessment Verification Test
PCS		Panel Closure System
ROM	Run-of-Mine
ROMPCS	Run-of-Mine (Salt) Panel Closure System
ROR	Rest of the Repository
RPD		Radiation Protection Division
SDDI	Salt Defense Disposal Investigation
SDI	Salt Disposal Investigation
SEN		Sensitivity 
SNL	Sandia National Laboratories
TDEM		Time Domain Electromagnetics
TRU		Transuranic
TSD	Technical Support Document
WIPP		Waste Isolation Pilot Plant
1.0 INTRODUCTION

The Waste Isolation Pilot Plant (WIPP) repository, located in deeply buried deposits of bedded salt in the Salado Formation in southeastern New Mexico, has been developed by the Department of Energy (the DOE) for the permanent disposal of transuranic (TRU) waste. The Environmental Protection Agency (EPA or the Agency) regulates containment of TRU waste in accordance with EPA's radioactive waste disposal standards at Code of Federal Regulations (CFR) Title 40, Parts 191 and 194. The WIPP was first certified by the EPA as complying with these regulations and approved for TRU waste disposal in 1998. The regulations require recertification of WIPP at five year intervals after the first waste shipment in 1999, with the most recent recertification occurring in 2014. EPA's decision to recertify WIPP is based in part on the results of an assessment of the projected ability of the facility to meet the Agency's waste isolation standards over the 10,000-year post-closure regulatory time frame. The ability to meet the numerical standards is determined by the results of modeling conducted for the DOE by Sandia National Laboratories (SNL). This modeling simulates the repository's future performance in a process called Performance Assessment (PA). The most recent assessment was included in the DOE's 2014 Compliance Recertification Application (CRA) and is called the CRA-2014 PA. 

Upon submittal by the DOE, EPA first reviews each CRA for completeness. Part of the DOE's responsibility is to update the PA to conform with recent changes in conditions that may affect post-closure repository performance such as the amount, types and properties of the waste and the rates of exploratory drilling for oil and gas. The the DOE is also responsible for updating the PA to conform with other changes including revisions to repository design, updated scientific information related to input parameters, and advances in science. A change in repository design occurred in 2012 in response to the DOE request to eliminate the requirement that the waste panel closures need to consist of a concrete monolith and explosion wall. The DOE proposed replacing the required design with one consisting of run-of-mine (ROM) salt. EPA's approval of that change was separately granted in 2012. Relevant additional scientific data and technical information related to repository performance include advances in understanding of the creep closure characteristics of salt, chemical conditions in the repository environment, as well as in the geochemical reactions and degradation processes of waste in WIPP's post-closure environment. 

EPA has identified several parameter updates based on recent experimental data that were not considered in the DOE's CRA-2014 PA that could affect repository performance. These updates included changes in geochemistry and in the creep characteristics of salt. At EPA's request, the DOE conducted four sensitivity studies to determine the effects that such changes would have on PA results and whether those changes had the potential to materially affect compliance with EPA's regulatory standards. This report documents EPA's review of the sensitivity study results and the effect of those results on compliance. 

During EPA's review of the CRA-2014, events associated with the February 2014 repository fire and deflagration event resulting in a small radionuclide release have closed portions of the underground facility. Since that time the DOE has determined to reconfigure or redesign certain areas of the repository in order to mitigate the effects of these two incidents. These changes may include revised panel closure designs and a revised waste panel layout. These sensitivity studies were designed and conducted at a time when the full effect of those closures on future repository operations was not known. The EPA deems that the sensitivity studies results remain relevant to future repository performance because they address fundamental processes such as the creep closure characteristics of Salado halite, the characteristics of waste degradation, the repository chemical environment, and the probability of Castile brine encounters that remain common to all WIPP repository designs. The results of the studies show that the WIPP remains in compliance despite the changes in PA parameters. However, EPA remains concerned about these changes and their effects on the Agency's lower probability compliance point (at the 0.001 probability compliance point).

2.0 SENSITIVITY STUDY OVERVIEW

Four sensitivity studies were conducted by SNL at EPA's request using modified parameter values identified by EPA. These studies were named CRA14_SEN1 through CRA14_SEN4 and will more simply be called SEN1 through SEN4 in this report. These studies evaluated how the WIPP release pathways were affected by the following modifications to WIPP PA.

 SEN1: Incorporate capillary effects and progressive creep closure of the WIPP operations and experimental area drifts.
 SEN2: Incorporate capillary effects and instantaneous creep closure of the WIPP operations and experimental area drifts.
 SEN3: Incorporate capillary effects and instantaneous creep closure of the WIPP operations, experimental, and panel closure area drifts.
 SEN4: Revise the probability of encountering a Castile brine reservoir beneath the repository; revise the uncertainty distributions for actinide solubility; correct the shear strength of degraded waste; revise the iron sulfidation reactions; use the corrected version of the DRSPALL code; and use the corrected panel closure length.  

The results of the sensitivity studies were compared with the results of the CRA-2014 PA (also called CRA14) to determine the effect of the changes on compliance.

3.0 SEN1 REVIEW

3.1 SEN1 Implementation

SEN1 was designed to determine the effects on fluid flow of modeling excavated drifts in the WIPP operations and experimental areas as progressively creep closing over time and approaching the properties of intact, undisturbed halite. This process is called "salt healing" and is expected to occur in a deeply buried salt formation such as the Salado. The ability of salt to creep and heal and encapsulate the waste was identified as a principal reason for selecting the WIPP's location (the DOE 1996, Chapter 1, p. 1). Creep closure of the drifts affects not only the drifts themselves but the consolidation and healing of the disturbed rock zone (DRZ) around the drifts. 

When the original WIPP certification was approved by the EPA in 1998, repository behavior and knowledge of salt creep and healing was not as advanced as it is today. At that time, the limitations in understanding were acknowledged and effectively addressed by incorporating parameter values and processes into WIPP performance modeling that would tend to overestimate radionuclide releases. Among these was the decision to not model creep closure of the operations and experimental areas but instead treat them as open drifts for the duration of the 10,000-year regulatory period. This was done to increase brine flow into the waste panels. The increased brine volume would increase brine saturation, increase the rate of waste degradation, and increase the resulting gas pressure buildup that drives several release pathways. However, modeling the operations and experimental areas as open drifts for 10,000 years also had the opposite effect of decreasing gas pressure by providing an artificially larger repository volume for gas storage. The reductions in calculated post-closure repository pressure accompanying enlargement of the experimental area for the Salt Disposal Investigations (SDI) provide an example of this effect (Camphouse et al. 2011, Section 6).

Since the initial WIPP certification, there have been advances in the understanding of salt creep and fluid flow as well as in the effects of assumptions that were once thought to unambiguously lead to overestimating WIPP releases. Because of these advances the EPA has revisited the effects of old assumptions to understand if they are aligned with current understanding of salt creep behavior and fluid flow and how these updates would affect repository performance.  The EPA understands that WIPP performance is complex, that the effects of changes to parameter values and creep closure processes on repository performance are often difficult to predict, and that because of these numerous complexities, the best way to determine the effects of those changes is through sensitivity studies. 

The principal parameters that control fluid flow in WIPP PA and were subject to change in the SEN1, SEN2, and SEN3 sensitivity studies are described below. 

Time Periods. Specific time periods were identified for modeling creep closure of WIPP drifts and adjacent DRZs. A five-year time period from -5 to 0 years is included in the PA model to establish initial repository conditions at closure (time zero).
Porosity and Permeability. These parameters were modeled in SEN1 as progressively decreasing over time due to salt creep, approaching the porosity and permeability of intact halite after 200 years. In SEN2 and SEN3 these parameters were modeled as approximating intact halite beginning at time zero. 
CAP_MOD. This is the capillary pressure model number. Model 1 is typically used when the threshold capillary pressure linear parameter (PCT_A) is set to zero, which deactivates the capillary pressure model. Model 1 activates the capillary pressure model when PCT_A is greater than zero but has no maximum capillary pressure. Model 2 activates the capillary pressure model in the same way as Model 1 but has a maximum capillary pressure (see PC_MAX, below).
COMP_RCK. This parameter is the bulk compressibility of the salt.
KPT. This parameter acts as a flag for using either constant or varying permeability in the threshold pressure calculation. When KPT is set to zero, the threshold capillary pressure is recalculated for each new material.
PC_MAX. This is the maximum value of capillary pressure, which is limited to avoid numerical problems in the BRAGFLO Salado flow model and is used with capillary pressure Model 2. Model 1 has no maximum capillary pressure but this is of no concern if the capillary pressure model is deactivated.
PCT_A. This is the threshold capillary pressure linear parameter. Setting this parameter to zero sets the threshold capillary pressure to zero and deactivates the capillary pressure model. Threshold capillary pressure can be zero for open fractures but not for porous media.
PCT_EXP. This is the threshold capillary pressure exponential parameter. This parameter is typically set to zero when PCT_A is set to zero. Setting both PCT_A and PCT_EXP to nonzero values activates the capillary pressure model.
PO_MIN. This is the minimum allowable brine pressure. This parameter is only used in capillary pressure Model 3 but is a required nonzero entry for all capillary pressure models.
PORE_DIS. This is the Brooks-Corey pore distribution parameter. This parameter is needed for relative permeability Model 4 which is required for the modified Brooks-Corey model that is used for all WIPP salt modeling.
RELP_MOD. This is the relative permeability model number. Relative permeability calculations for WIPP salt use the modified Brooks-Corey model, which corresponds to relative permeability Model 4.
SAT_IBRN. This is the initial brine saturation of the salt. This parameter should be greater than the residual brine saturation to avoid numerical problems in BRAGFLO.
SAT_RBRN. This is the residual brine saturation of the salt. Residual brine saturation is the brine saturation at which brine flow ceases due to capillary tension holding the small residual brine content in place and only gas can flow. Residual brine saturation can approach zero for open fractures where capillary effects may be small but should be greater than zero in a porous medium.
SAT_RGAS. This is the residual gas saturation. Residual gas saturation is the gas saturation at which gas flow ceases because the brine saturation has become so high that no continuous open pathways for gas remain. Residual gas saturation can approach zero for open fractures because there could be continuous pathways for gas at very high brine saturations due to a lack of capillary forces holding brine. Residual gas saturation should be greater than zero in a porous medium.
For the SEN1 sensitivity study, the EPA requested SNL to make the parameter value changes identified in Tables 1 and 2. These changes affected both the operations and experimental area drifts and the DRZs adjacent to those drifts. They were made to simulate progressive creep closure of the drifts as well as progressive healing of the DRZs to near undisturbed conditions and include an activated capillary pressure model. A comprehensive discussion of the technical basis for these changes is presented in EPA (2017a). 

3.2 SEN1 Results

In previous performance assessments SNL found that the computer model BRAGFLO, which simulates fluid flow in the Salado, would not function properly when permeability was progressively changed and the capillary pressure model was activated. SNL's resolution of this problem was to turn off (deactivate) the capillary pressure model. The Agency believed this problem was caused by setting an initial brine saturation that was lower than the residual brine saturation, as discussed in the Agency's review of the DOE's 2012 PCS PA (EPA 2013). The Agency believed that the problem had been resolved by setting the initial brine saturation to a high value such as 1.0. However, when this remedy was tested by SNL in SEN1, the BRAGFLO model would converge but resulted in non-physical fluctuations in pressure. Therefore, the model output was suspect and no results were produced. The DOE informed EPA that the BRAGFLO model would only properly converge under the following conditions.

 If the permeability is varied over time, the capillary pressure model must be turned off.
 If the permeability is constant over time, the capillary pressure model can be turned on.

In order to test the effects of capillary pressure in a two-phase flow regime, the DOE proposed running EPA's sensitivity study by adopting the following approach: a) eliminate the progressive salt healing process; and b) assume the operations and experimental area drifts, and the adjacent DRZs, have already healed to properties approaching intact halite at the beginning of the analysis. This approach avoids progressive changes in permeability and allows the BRAGFLO model to properly converge when capillary effects are included. 

3.3 EPA SEN1 Conclusions

Considering that including, rather than ignoring, capillary effects would provide a more accurate analysis and that the initial 200-year period of progressive porosity and permeability decrease is relatively short compared with the 10,000-year regulatory period, the Agency accepted the DOE's proposal and implemented it in the SEN2 sensitivity study.

4.0 SEN2 REVIEW

4.1 SEN2 Implementation

BRAGFLO convergence issues identified during the SEN1 study were used as a starting point for the SEN2 study.  For the SEN2 study, instead of trying to gradually reduce permeability and increase capillary pressure with time, it was assumed that the operations and experimental room open drifts and adjoining DRZs would be healed at the time of repository closure. SEN2 was designed to determine the impact on repository performance when assuming these areas are fully healed and two-phase flow parameters are adopted. These areas were therefore modeled by adopting properties that approach those of intact, undisturbed halite. The parameter values adopted for SEN2 are shown in Tables 3 and 4. 

The types of fluid flow parameters considered for the SEN2 study are the same as those described above for SEN1 with the following exceptions. The permeability of the DRZ was reduced to better reflect the more complete healing expected in the DRZ due to the essentially complete re-mating of tensile fractures. Also, the initial brine saturation was slightly reduced to better reflect initial conditions. Residual brine and gas saturations were also increased to better reflect healed conditions. 

SEN2 implementation is described by Day (2016, Section 2). The BRAGFLO model was modified for the SEN2 study to incorporate the parameter values listed in Tables 3 and 4.  The model was run to simulate the undisturbed repository (scenario S1-BF) as well as the five standard disturbed scenarios used to represent inadvertent human intrusion after repository closure. The disturbed scenarios consist of two single future drilling intrusions into the repository that also intersect a Castile brine reservoir (one at 350 years and one at 1,000 years after closure: scenarios S2-BF and S3-BF); two single future drilling intrusions into the repository that do not intersect a Castile brine reservoir (one at 350 years and one at 1,000 years after closure: scenarios S4-BF and S5-BF); and a double intrusion (one at 1,000 years after closure and a second at 2,000 years after closure: scenario S6-BF). The first borehole in scenario S6-BF does not intersect a Castile brine reservoir and the second penetrates the same waste panel as the first and does intersect a brine reservoir. 

Only one of the ten WIPP waste panels is specifically modeled in BRAGFLO to represent the consequence of intersecting boreholes. That panel is referred to as "the waste panel" in this report and represents the southwestern panel, called Panel 5. This panel was selected for specific representation because it is located in the most southerly down dip area of the repository and accumulation of brine seepage into the repository would be maximized in that panel. Because this panel tends to accumulate the most brine, the conditions in this panel are most likely to activate the important direct brine release (DBR) pathway that is most sensitive to both waste panel pressure and brine volume. 

The standard three replicates were run in SEN2, each consisting of 100 vectors. All modified parameters that are defined in Tables 3 and 4 are the same as the sampled or calculated values for intact halite used, vector by vector, in the CRA-2014 PA. This was done to facilitate comparison with CRA-2014 PA results and thereby determine the sensitivity of those results to the changes in parameter values and the incorporation of creep closure. 

4.2 SEN2 Results

SNL's review of SEN2 results is presented in Day (2016, Section 4). Day (2016) presents these results as averages across all three replicates. The SEN2 parameter value changes, identified in Tables 3 and 4, restricted brine and gas flow in the operations and experimental areas, and also essentially eliminated their ability to store gas. When a Castile brine reservoir was not intersected, this combination of effects decreased average brine flow into the waste panel and increased average brine pressure in the panel, as shown in Figures 1 and 2. When a Castile brine reservoir was intersected, these effects tended to be overwhelmed by the rapid flow of large volumes of pressurized Castile brine into the waste panel, as shown in Figures 3 and 4. Although waste panel pressure was slightly higher in SEN2 when a brine reservoir was intersected, cumulative brine flow into the panel was unchanged.

Cuttings and cavings are the major pathways for high probability, high consequence releases but these do not depend on repository brine volume or pressure and, as expected, were not affected by the SEN2 changes. Of the pathways that do depend on brine volume and/or pressure, DBRs are the most important, followed by spallings releases and brine flow up the borehole and through the Culebra dolomite. The remaining release pathways are brine flow up the shaft and through the anhydrite interbeds but these are small and provide negligible contributions to total releases. The following paragraphs describe the principal effects of the SEN2 changes on the release pathways that are sensitive to brine volume and/or pressure.   

Direct Brine Releases. DBRs are a function of both brine volume and pressure, and are the major source of low probability but high consequence repository releases. Waste panel conditions of high brine saturation and high pressure are most conducive to the occurrence of DBRs and those conditions are most often associated with Castile brine reservoir encounters. The maximum average cumulative brine inflow into the waste panel in SEN2 amounted to about 4,000 m[3] for the undisturbed case, about 6,000 m[3] when a borehole intersects a waste panel but not a brine reservoir, and about 26,000 m[3] when a borehole intersects both a waste panel and also a brine reservoir (Day 2016, Figures 4-53, 4-54, and 4-55). Similarly, the maximum average pressure in SEN2 in the waste panel is about 7 MPa for the undisturbed case and about 5 MPa when a borehole intersects a waste panel but not a brine reservoir (Day 2016, Figures 4-17 and 4-19). As shown in Figure 3, when a borehole intersects both a waste panel and also a brine reservoir, the average brine pressure peaks at about 11 MPa and then slowly declines to about 8 MPa in both SEN2 and CRA-2014.

A minimum brine pressure of about 8 MPa is required for a DBR to be initiated (the DOE 2014, Appendix PA, Section PA-4.7.1). The average calculated pressure reported by Day (2016) for an intrusion scenario can be compared with the 8 MPa threshold pressure to determine whether or not this threshold is likely to be exceeded in any individual vector within that scenario. Some individual vectors will have calculated pressures less than the average and some will have pressures that are greater. If the average calculated pressure is greater than 8 MPa, the calculated pressure in most individual vectors is also likely to be greater than 8 MPa. This minimum pressure requirement is therefore likely to be met in more than half of the vectors when the average calculated pressure is greater than 8 MPa. Average calculated pressures greater than 8 MPa therefore become more favorable to the initiation of DBRs. Individual vectors with pressures greater than 8 MPa do occur when the average calculated pressure is less than 8 MPa, but their occurrence becomes less likely as the average drops and the occurrence of conditions supporting the initiation of DBRs become less favorable. 

In addition to the requirement that the pressure exceed 8 MPa, the calculated brine saturation must be greater than the sampled residual brine saturation for a DBR to occur because, as described in Section 3.1, brine cannot flow when the actual saturation is less than the residual saturation. The residual brine saturation in the waste panel was sampled from a uniform distribution ranging from 0.0 to 0.552 in both SEN2 and the CRA-2014 PA independent of the calculated saturation (the DOE 2014, Appendix PA, Section PA-4.7.1). The mean sampled residual brine saturation is approximately 0.3. If the average calculated saturation for a scenario is less than this mean, the average sampled residual saturation is likely to be greater than the calculated saturation and therefore unfavorable to the initiation of DBRs. Again, calculated saturations for individual vectors within a scenario can be either greater or less than the mean and some can therefore be greater than the sampled residual saturation even when the average calculated saturation is less than 0.3. However, in evaluating trends, as the calculated saturation drops to values lower than 0.3 it becomes less likely to be high enough to support a DBR.

Although average pressures in the waste panel increased slightly in SEN2 as compared with CRA-2014, the results in Figures 1 and 3 indicate that average pressures exceeding the 8 MPa DBR threshold are only present in both SEN2 and CRA-2014 when a Castile brine reservoir is encountered. Average brine saturations in the waste panel decreased slightly in SEN2 consistent with the increases in repository pressure but remained close to the saturations in CRA-2014. The average brine saturation in the waste panel in both SEN2 and CRA-2014 was about 0.25 for the undisturbed case, about 0.4 to 0.5 when a borehole intersects a waste panel but not a brine reservoir, and about 0.9 when a borehole intersects both a waste panel and a brine reservoir (Day 2016, Figures 4-84, 4-85, and 4-86). Both brine saturation and brine pressure would therefore be typically too low in the undisturbed case to support a DBR but both would be practically assured of being high enough to support a DBR when a Castile brine reservoir is encountered. In the intermediate case of a borehole intersecting a waste panel but not a brine reservoir, the average brine saturation is high enough but the average pressure is too low for a DBR to be likely. 

Figure 4 shows that, compared with the CRA-2014 results, the average brine inflows into the waste panel are essentially unchanged in SEN2 when a Castile brine reservoir is encountered, indicating that brine inflows are controlled by the pressure and volume of the brine reservoir and the small pressure increase in SEN2 had essentially no effect on the volume of those inflows.

The pattern of conditions favorable to DBR releases in SEN2 was also observed in the CRA-2014 results. Conditions favorable to DBRs in both SEN2 and CRA-2014 only routinely occurred when a Castile brine reservoir was encountered. Although pressures tended to increase and brine flow tended to decrease in SEN2, the differences between the SEN2 and the CRA-2014 results were not enough to substantially change this pattern. 

Based on the foregoing observations, DBRs are most likely to occur when Castile brine reservoirs are encountered and are unlikely to occur when brine reservoirs are not encountered. Because the SEN2 parameter changes had little effect on waste panel conditions when brine reservoirs were encountered, little change in DBRs occurred. This result is illustrated in Figure 5 where no differences in mean DBR releases are evident between CRA-2014 and SEN2.

Spallings Releases. Spallings was the only pathway that showed a noticeable increase in releases in SEN2. Unlike DBRs, spallings releases are a function of repository pressure and not brine saturation. For a spallings release to be initiated, the repository pressure must be at least 2 MPa greater than the well bottom pressure of approximately 8 MPa (Kicker et al. 2016, Sections 5.4.6 and 6.1.2.1). Spallings releases can therefore occur in both the first intrusion and in subsequent intrusions when the pressure is high. As shown in Figure 6, spallings releases did increase in SEN2 as compared with the CRA-2014 results but total spallings releases remained small compared with cuttings, cavings, and DBR releases. Spallings releases therefore did not materially contribute to total repository releases in either SEN2 or CRA-2014.

Other Releases. Brine flows up the repository shaft decreased in SEN2 as compared to CRA-2014 PA results due to restricted flow within the operations and experimental areas. Releases through the Culebra were not affected by the changes made for SEN2 and, as expected, cuttings and cavings releases were unchanged. A comparison of releases by the major release pathways for SEN2 is presented in Figure 7.

Total Releases. The effects of the SEN2 changes on total mean high-probability (P(R) > 0.1) releases were insignificant because those releases are dominated by cuttings and cavings which are not affected by repository brine saturation and pressure. Total mean low-probability (P(R) > 0.001) releases, primarily DBR, spallings and Culebra releases that are affected by repository brine saturation and/or pressure, increased by about 4%. The 95% confidence level on the mean reduced by about 20% (Day 2016, Section 5). Total mean releases for SEN2 and CRA-2014 are plotted on Figure 8. At the scale of the figure, the differences between the two are nearly indistinguishable.  Individual total normalized release vectors for the three replicates of the SEN2 study are plotted on Figure 9. As can be seen from the figure, no individual vectors exceeded EPA's release limits.

4.3 EPA SEN2 Conclusions

The EPA concludes that the SEN2 parameter changes tended to increase waste panel pressure and decrease brine inflow and saturation in scenarios that do not involve Castile brine encounters but these changes were too small to significantly affect releases. In scenarios where Castile brine encounters do occur the effect of the SEN2 changes was also small because any increase in pressure or decrease in saturation due to the parameter changes were swamped by the pressure and saturation increases due to the inflowing Castile brine. Because total low probability repository releases are most affected by Castile brine encounters and total high probability releases were not affected by the SEN2 parameter changes, modeling creep closure and healing of the operations and experimental areas of the repository was shown to have little effect on the prediction of total releases from the repository. EPA's WIPP release limits were therefore not exceeded by mean total releases, the 95% confidence level in that mean, or by any individual vectors in SEN2.

5.0 SEN3 REVIEW

5.1 SEN3 Implementation

The objective of the SEN3 study was to simulate repository behavior that includes capillary effects as well as creep closure and healing in all non-waste area drifts and surrounding DRZs. The SEN3 study was therefore essentially the same as SEN2 but with the addition of incorporating creep closure and healing of ROM salt panel closures in the waste panel access drifts and adjacent DRZs. The changes made in SEN3 are significant in that a comparison with CRA-2014 results provides insight into the effects on repository behavior of two alternative performance modeling approaches for non-waste areas: the intentionally more conservative, open drift approach used the CRA-2014 PA, and the more realistic, closed drift approach used in SEN3. As discussed in Section 3.1, the intent of the open drift approach was to enhance brine movement throughout the repository and increase brine saturation in the waste panels, thereby increasing waste degradation rates and gas pressure buildup. As a result, the open drift approach was expected to conservatively increase calculated repository releases beyond what would actually occur. To achieve this, modeling of the operations and experimental area drifts in CRA-2014 intentionally left out creep closure for the duration of the regulatory time frame. Additionally, it did not include complete healing of the ROM salt panel closures and adjacent DRZs and it ignored capillary effects which could further limit brine and gas flow. 

At the time the CRA-2014 was submitted the waste panel access drifts were to be backfilled with run-of-mine (ROM) salt and the majority of the operations and experimental area drifts were allowed to close naturally through salt creep without backfill. The Agency has not been notified whether the waste panel closure design will be modified as a result of the February 2014 repository fire and radionuclide release. The ROM salt panel closure design submitted in the CRA-2014 consisted of salt aggregate produced as a byproduct of waste panel mining. A description of the ROM salt panel closure design is provided in EPA (2013). Over time the salt aggregate will be compressed and consolidated by creep closure of the surrounding halite and will eventually return to a state approaching that of intact Salado halite. 

The operations and experimental area drifts will progressively close over time after they are no longer maintained and will eventually be filled with rubble primarily from roof collapse. The rubble will be compressed and consolidated by continuing creep of the surrounding halite and will also eventually return to a state approaching that of intact Salado halite. Closure of the waste panel access drifts may be faster than the operations and experimental area drifts due to the presence of an initial backfill. Creep closure of all non-waste drifts is expected to be complete within several hundred years. As previously noted, creep closure of these drifts affects not only the drifts themselves but promotes healing of the surrounding DRZ which will also eventually acquire properties similar to those of intact halite.
The Agency's initial plan for SEN3 was to model the progressive creep closure of the backfilled waste panel access drifts and surrounding DRZs over time using an approach similar to that originally proposed for the operations and experimental area drifts in SEN1. This approach would have simulated the decreases in porosity and permeability as the access drifts and surrounding DRZs were compressed and consolidated by creep flow of the surrounding halite. The theoretical basis for this process and its timing are described in EPA (2017a) and the initially proposed modeling parameters are summarized in Tables 5 and 6. Closure of the access drifts was to be modeled in two 50-year stages to allow for the more rapid consolidation of the ROM salt backfill than the DRZ. Consolidation is expected to occur relatively quickly under WIPP conditions and the properties of both the ROM salt and DRZ are expected to approach those of intact halite within about 100 years. By comparison, approximately 200 years were allowed for the operations and experimental area drifts to reach a similarly healed state because backfilling is not planned in those drifts at the time of repository closure. This difference is not expected to be significant as compared with the 10,000-year regulatory period.

EPA would have preferred to model progressive rather than instantaneous drift closure but the same modeling constraints encountered in the SEN1 study persisted and the BRAGFLO model could not simulate both the progressive closure and the capillary effects inherent in the originally proposed parameter values. The assumptions of instantaneous consolidation of the ROM salt panel closures as well as immediate creep closure and healing of the operations and experimental areas were therefore accepted by the Agency for SEN3 and the parameter values in Tables 5 and 6 were not used. 

SEN3 also modeled healing in the DRZ adjacent to the panel closures and operations and experimental area drifts at the time of repository closure. The parameter values adopted for the operations and experimental areas in SEN3 are the same as for SEN2 and are shown in Tables 3 and 4. The final parameter values adopted for the panel closures in SEN3 are shown in Tables 7 and 8. Because creep closure of the drifts and healing of the DRZ will occur relatively quickly, the Agency concludes that the assumption of immediate closure is acceptable. 

The types of fluid flow parameters considered for SEN3 are the same as those considered for SEN1 and SEN2. The parameter values are also the same with the principal exception that the ROM salt is expected to heal to the porosity and permeability that are closer to the values for intact halite, which are slightly lower than the end-point values for the rockfall rubble in the operations and experimental area drifts. This conclusion is supported by recent laboratory studies of healing aggregate salt backfill described in EPA (2017a). Residual brine and gas saturations in the panel closures were also modified based on recent laboratory studies described in EPA (2017a). In addition to the changes in parameter values in Tables 7 and 8, EPA also requested the DOE to simulate the two sets of panel closures planned between the operations area and Waste Panel 10 instead of the single set that was previously being modeled. This was accomplished by doubling the length of those panel closures in the BRAGFLO grid.

SEN3 implementation is described by Day and Zeitler (2016, Section 2). The BRAGFLO model used for SEN2 was modified for SEN3 to incorporate the parameter values listed in Tables 7 and 8. The model was run following essentially the same set of protocols that were used in the SEN2 study and described in Section 4. The analysis again simulated the undisturbed repository as well as the five standard disturbance scenarios used to represent inadvertent human intrusion after repository closure. The standard three replicates were also run. All modified parameters that are defined in Tables 7 and 8 are the same as the sampled values for intact halite in CRA-2014.  As with SEN2, these values correspond to the sampled and/or calculated values in the CRA-2014 PA on a vector by vector basis. This was done to facilitate comparison with the CRA-2014 PA results and determine the sensitivity of those results to the changes in parameter values. 

5.2 SEN3 Results

SNL's review of SEN3 results is presented in Day and Zeitler (2016, Section 4). Day and Zeitler (2016) present those results as averages across all three replicates. The SEN3 parameter value changes restricted brine and gas flow through the waste panel access drifts as well as in the operations and experimental areas and essentially isolated the waste panels from one another as well as from the operations and experimental areas. As compared with SEN2, the additional SEN3 parameter value changes generally resulted in increased pressures and decreased brine saturations in the waste areas, and decreased brine and gas flows across all panel closures.

The trends in SEN3 results for the various release pathways were similar to those for SEN2. Cuttings and cavings do not depend on repository brine volume or pressure and were again not affected by the SEN3 parameter value changes. Of the pathways that do depend on brine volume and/or pressure, DBR releases were only moderately increased but spallings releases increased as compared with both SEN2 and CRA-2014. Releases from brine flow through the Culebra were again unchanged, and brine flow up the shaft and through the anhydrite interbeds were again small and did not materially contribute to total releases. 

The Agency's review of SEN3 results is documented in greater detail than for the SEN2 results because the SEN3 study provides a more comprehensive analysis by incorporating creep closure of ROM salt panel closures in the waste panel access drifts as well as the operations and experimental area drifts. Emphasis in this review has been placed on the effects of the SEN3 changes on DBRs because of the important contributions of that release pathway to total repository releases. 

Direct Brine Releases  -  Scenario S1-BF: Undisturbed Repository. The undisturbed condition describes the state of the repository during the time before the first borehole intersection. As previously described, DBRs occur only when the pressure of the intersected waste panel is above 8 MPa and the brine saturation in the waste panel is above the sampled residual saturation, which averages about 0.3. Again, DBRs are less likely to be initiated in a given scenario when the average pressure is below 8 MPa and/or the average brine saturation is below 0.3. 

Comparisons of average pressure, cumulative brine inflow, and residual saturations in the single modeled waste panel for undisturbed conditions in SEN3 and CRA-2014 are provided in Figures 10, 11, and 12. SEN2 results are also shown on these figures but the emphasis in this discussion is on the more comprehensive SEN3 results. Figure 10 shows that the average waste panel pressure increased from about 6.5 MPa in CRA-2014 to 9.5 MPa in SEN3. Figure 11 shows that average cumulative brine inflow into the waste panel decreased from over 4,000 m[3] in CRA-2014 to about 3,000 m[3] in SEN3 due to the increased pressure. Figure 12 shows that average brine saturation decreased from about 0.3 in CRA-2014 to 0.1 in SEN3 due to the increased pressure and decreased brine volume. The first borehole penetration into the repository will encounter undisturbed conditions. These results show that, on the average, in CRA-2014 the calculated brine saturation in the single modeled waste panel is equal to the average residual saturation but the pressure is too low to initiate a DBR. Also on the average, in SEN3 the pressure is sufficient to initiate a DBR but the calculated brine saturation is well below the average sampled residual saturation, making it unlikely that the calculated saturation would be high enough in an individual vector to support a DBR.

DBRs can also occur from borehole penetrations of waste panels to the north of the single modeled southwestern waste panel. These panels are grouped into the north and south rest of the repository (ROR) and pressure and brine saturation conditions in those areas are computed in BRAGFLO. Average pressures and saturations for the undisturbed repository are shown in Day and Zeitler (2016, Figures 4-9 and 4-76) for the north ROR and Day and Zeitler (2016, Figures 4-13 and 4-80) for the south ROR. The parameter changes in SEN3 did not substantially change either the pressures or brine saturations in these areas. In the north ROR the average pressures were typically about 5.5 MPa in both SEN3 and CRA-2014 and average brine saturations were typically about 0.1 in both analyses. In the south ROR the average pressures were typically about 6 MPa in both SEN3 and CRA-2014 and average brine saturations were again typically about 0.1 in both analyses. On the average, the pressures in the RORs were too low to support a DBR and the brine saturations were well below the average sampled residual saturation. DBRs are therefore unlikely to occur in the RORs under these conditions.

It can be concluded that in both SEN3 and CRA-2014, DBRs would be rare in the first penetration of a previously undisturbed repository and would only occur in those few vectors where both brine pressure and saturation are sufficiently high.

Direct Brine Releases  -  Scenario S4-BF: Borehole Penetrates the Repository at 350 Years but not a Castile Brine Reservoir. Penetration of the repository but not a brine reservoir is the most common disturbed repository condition because on average fewer than half the borehole intersections of the waste area result in a Castile brine encounter. Borehole intrusions of this type are modeled in WIPP PA for borehole intersections of the waste panel at 350 and 1,000 years after repository closure but only the results at 350 years are presented by Day and Zeitler (2016, p. 27) because the results at 1,000 years are similar.  

Comparisons of average pressure, cumulative brine inflow, and residual saturations in the single modeled waste panel for scenario S4-BF in SEN3 and CRA-2014 are provided in Figures 13, 14, and 15. Figure 13 shows that the average waste panel pressure increased slightly from about 5 MPa in CRA-2014 to 5.5 MPa in SEN3. Figure 14 shows that average cumulative brine inflow into the waste panel decreased from about 6,500 m[3] in CRA-2014 to about 5,500 m[3] in SEN3 due to the increased pressure. Figure 15 shows that average brine saturation decreased from about 0.45 in CRA-2014 to 0.4 in SEN3 due to the increased pressure and decreased brine inflow. These results show that, on the average, in both SEN3 and CRA-2014 the brine saturation in the single modeled waste panel is above the sample average but the average pressure is below the threshold needed to support a DBR. 

Direct brine releases are also assumed to occur in WIPP PA in boreholes penetrating the north and south ROR, however the BRAGFLO grid is configured with only one borehole and that borehole passes through the single modeled waste panel described above. As a result, the effect on repository conditions of that borehole penetration is calculated only for that single modeled waste panel. The conditions in the rest of the repository essentially reflect the undisturbed state as modified by whatever changes in conditions, such as pressure and brine flow, emanate through the panel closure system, the DRZ, and the anhydrite interbeds into the RORs from the single modeled waste panel. Because these materials have relatively low permeability, the effect of borehole penetrations in the single modeled waste panel on conditions in the north and south RORs would be expected to be small. Average pressures and saturations for S4-BF penetrations are shown in Day and Zeitler (2016, Figures 4-11 and 4-78) for the north ROR and in Day and Zeitler (2016, Figures 4-15 and 4-82) for the south ROR. The parameter changes in SEN3 increased the pressure by about 1 MPa as compared with CRA-2014 but did not substantially change the brine saturations in these areas. In both the north and south RORs the average pressures increased from about 5 MPa to 6 MPa and the average brine saturations were typically about 0.1 in both analyses. These saturations are lower than the S4-BF saturations of 0.4 to 0.45 in the single modeled waste panel, reflecting the lack of direct modeling of borehole intrusions in the RORs. On the average, both pressures and brine saturations in the RORs are again too low to make DBRs a likely occurrence.

It can again be concluded that in both SEN3 and CRA-2014, DBRs would be unlikely in both the waste panel and the RORs in an S4-BF scenario where a Castile brine reservoir is not intersected and would only occur in those few vectors where both brine pressure and saturation are sufficiently high.

Direct Brine Releases  -  Scenario S2-BF: Borehole Penetrates the Repository at 350 years and also a Castile Brine Reservoir. This is a less common disturbed repository condition because on the average fewer than half the borehole intersections of the waste area encounter Castile brine. Borehole intrusions of this type are again modeled for borehole intersections at 350 and 1,000 years after repository closure but only the results at 350 years are presented by Day and Zeitler (2016, p. 27) because the results at 1,000 years are similar.  

Comparisons of average pressure, cumulative brine inflow, and residual saturations in the single modeled waste panel for disturbed conditions in SEN3 and CRA-2014 when a Castile brine reservoir is encountered are provided in Figures 16, 17, and 18. Figure 16 shows that in CRA-2014 the average waste panel pressure peaks early in the regulatory period at about 11 MPa and then slowly decreases to about 8 MPa at 10,000 years. Figure 16 also shows that in SEN3 the average waste panel pressure peaks at the same time to slightly above 11 MPa and then slowly decreases to about 9 MPa at 10,000 years. The average cumulative brine inflow in Figure 17 shows dramatic increases as compared with the two previous scenarios. The average cumulative brine inflow into the waste panel increases from typically less than 6,500 m[3] to about 26,000 m[3] when a brine reservoir is encountered and the inflow volume is the same in both SEN3 and CRA-2014. Figure 18 shows that average brine saturation jumps to nearly 100% (1.0) due to inflowing brine in both SEN3 and CRA-2014 when a Castile brine reservoir is encountered and then decreases to about 0.9 in CRA-2014 and 0.8 in SEN3. The greater decrease in SEN3 is attributed to the increased pressure in that analysis. These results show that in SEN3 and CRA-2014 the average brine saturation and pressure are both high enough to make DBRs a likely occurrence. 

Average pressures and saturations in the both the south and north ROR for the S2-BF scenario are shown in Day and Zeitler (2016, Figures 4-10 and 4-77) for the north ROR and in Day and Zeitler (2016, Figures 4-14 and 4-81) for the south ROR. The pressures in both areas were typically about 6 MPa but were surprisingly slightly lower in SEN3 than in CRA-2014. Although pressures generally increased in SEN3, the Agency attributes this unusual decrease to the fact that the pressure increase from the inflowing brine only occurs in the modeled waste panel and the transmission of that pressure through the tighter panel closure system in SEN3 and into the south and north ROR was smaller in SEN3 than in CRA-2014. The average brine saturations were typically about 0.1 in both the south and north ROR in both analyses. On the average, both pressures and brine saturations were again typically too low in the RORs to make DBRs a likely occurrence.

Because the Castile brine reservoir is below the repository horizon, in the S2-BF scenario the intersecting borehole encounters the repository first and then encounters the brine reservoir later. Whether or not a DBR is initiated in the first borehole that intersects a brine reservoir depends on the prior repository condition, which can either be an S1-BF, S4-BF, or S5-BF. The next borehole that intersects the same panel following the first S2-BF intrusion would encounter the pressures and brine saturations depicted in Figures 16 and 18. It can be concluded that in both SEN3 and CRA-2014, DBRs would be common in a borehole that intersected a waste panel that had a previous Castile brine encounter because both brine pressure and saturation would, on the average, be sufficiently high.

Direct Brine Releases - Scenario S6-BF: A Double Borehole Intrusion; One at 1,000 Years and a Second at 2,000 Years. The first borehole in this scenario is modeled as an S5-BF that does not intersect a Castile brine reservoir while the second borehole is modeled to intersect a brine reservoir 1,000 years later. This scenario is intended to simulate the possibility that Castile brine flooding a waste panel could flow through the waste and exit the panel through a previously drilled borehole to the accessible environment. 

Comparisons of average pressure, cumulative brine inflow, and residual saturation in the single modeled waste panel for this double borehole intrusion scenario are provided in Figures 19, 20, and 21. Figure 19 shows that in this scenario the pressure abruptly increases to about 9 MPa when the brine reservoir is hit at 2,000 years in both studies and then subsequently decreases to about 7 MPa in CRA-2014 and 8.5 MPa in SEN3. Figure 20 shows that the average cumulative brine inflow amounts to about 20,000 m[3] in CRA-2014 and is decreased slightly to 19,000 m[3] in SEN3. This decrease is attributed to the higher pressure in SEN3. Figure 21 shows that average brine saturation jumps to about 80% (0.8) due to inflowing brine in both SEN3 and CRA-2014 when a Castile brine reservoir is encountered. Following this increase, the average saturation remains at about 0.8 in CRA-2014 and decreases to about 0.7 in SEN3. The decrease in SEN3 is again attributed to the increased pressure in that analysis. The patterns of waste panel pressure and saturation change after the brine reservoir is encountered at 2,000 years are similar to the patterns observed when a reservoir is encountered at 350 years in the S2-BF scenario, with the differences attributed primarily to the moderating effect of pressure leakage through the first borehole intrusion and the shorter regulatory time duration following the second intrusion. These results indicate that while the average pressure drops to below the 8 MPa DBR threshold in CRA-2014, it remains above this threshold in SEN3 and average brine saturations remain high in both analyses. These results suggest that in the S6-BF scenario, DBRs may be more frequent in SEN3 than in CRA-2014. 

Average pressures and saturations in the RORs for the S6-BF scenario are shown in Day and Zeitler (2016, Figures 4-12 and 4-79) for the north ROR and in Day and Zeitler (2016, Figures 4-16 and 4-83) for the south ROR. The pressures in both areas were typically about 6 MPa but the parameter changes in SEN3 decreased the pressure slightly. This decrease is again attributed to the lower permeability of the waste panel closure system in SEN3 that reduced gas and brine flow out of the intruded, high pressure panel and into the RORs. The average brine saturations were typically about 0.1 in both areas and both analyses. On the average, both pressures and brine saturations were again typically too low in the RORs to make DBRs a likely occurrence.

Direct Brine Releases  -  EPA Conclusions. Average conditions in the waste panel and the south and north ROR are indicative of whether, on the average, a DBR would occur in response to a borehole penetration. DBRs are more likely to occur if the average pressure is above the threshold value of about 8 MPa and the average calculated brine saturation is above the mean sampled residual brine saturation of about 0.3. If either of these averages is below the respective reference value, a DBR can only occur in the less likely event that an individual vector in the scenario has both a pressure above 8 MPa and a brine saturation above the residual saturation. 

The only scenarios where DBRs were found to likely occur involve a Castile brine encounter and the subsequent release of Castile brine into the waste panel. DBRs were found to be less likely to occur in the S1-BF undisturbed repository scenario, in the S4-BF and S5-BF disturbed scenarios involving a borehole that does not intersect a Castile brine reservoir, and in all scenarios involving borehole penetrations of the north and south RORs. DBRs were only found to likely occur in the S2-BF, S3-BF, and S6-BF scenarios that all involve Castile brine encounters. The results also indicate that repository conditions that influence pressure have a greater impact on releases than conditions that influence brine saturation. Although the average pressure was below the 8 MPa DBR threshold in CRA-2014 in the S6-BF scenario, it was above that threshold in SEN3 and average brine saturations remained high in both analyses. These results suggest that the generally higher pressures in SEN3 could result in an increase in DBRs as compared with CRA-2014. This conclusion is supported by the comparison of DBR results shown in Figure 22 

A graphical presentation of results for the single modeled waste panel is presented in Figure 23. This figure illustrates the following summary of results.
 Average pressures are higher and average brine saturations are lower in SEN3 than in CRA-2014.
 In SEN3:
 Average pressures are higher than the DBR threshold pressure of 8 MPa in three of the four scenarios, excluding only S4-BF.
 Average brine saturations are higher than the DBR reference residual saturation of 0.3 in three of the four scenarios, excluding only S1-BF.
 The combination of higher average pressure and higher average brine saturation needed to create likely conditions for a DBR is present only in scenarios S2-BF and S6-BF, both of which involve a Castile brine reservoir encounter.
 In CRA-2014:
 Average pressures are equal to or higher than the DBR threshold pressure of 8 MPa in only one of the four scenarios, S2-BF.
 Average brine saturations are equal to or higher than the DBR reference residual saturation of 0.3 in all of the four scenarios.
 The combination of higher pressure and higher brine saturation needed to create likely conditions for a DBR is present only in scenario S2-BF which involves a Castile brine reservoir encounter.
 For both SEN3 and CRA-2014, the pressure and saturation conditions needed to initiate a DBR are likely to occur only when a Castile brine reservoir is encountered.

Spallings Releases. As shown in Figure 24, spallings releases were greater in SEN3 than in SEN2 and both were greater than in CRA-2014. This is likely due to the greater pressures in the waste panel in all scenarios. However, spallings releases remained small compared with cuttings, cavings, and DBR releases. Therefore, the increase in spalling releases did not materially contribute to total repository releases.

Other Releases. Brine flows up the repository shaft decreased in SEN3 as compared to CRA-2014 PA due to restricted flow within the operations and experimental areas. Releases through the Culebra were not affected by the changes made for SEN3 and remained less than spallings releases except at very low probabilities. Cuttings and cavings releases were unchanged as expected. A comparison of releases for the major release pathways for SEN3 is presented in Figure 25.

Total Releases. The effects of the SEN3 changes on total mean high-probability (P(R) > 0.1) releases were insignificant because those releases are dominated by cuttings and cavings which are not affected by repository brine saturation and gas pressure. Total mean low-probability (P(R) > 0.001) releases increased by about 15% in SEN3 as compared with a 4% increase in SEN2. The increase at this low probability was due to an increase in DBR and spallings releases that are affected by repository brine saturation and/or gas pressure. The 95% confidence level on the mean was nearly unchanged (Day and Zeitler 2016, Section 5). Total releases for SEN2, SEN3 and CRA-2014 are plotted on Figure 26. At the scale of the figure, the increased releases under SEN3 are small but distinguishable and are attributed to increases in DBRs. A plot of individual total normalized release vectors for the three replicates of the SEN3 study is presented in Figure 27. As can be seen from the figure, no individual vector exceeded the Agency's release limits.

5.3 EPA SEN3 Conclusions

The EPA concludes that modeling creep closure and healing of the waste panel access drifts, the operations and experimental area drifts, and the DRZs associated with those drifts results in only a small increase in total releases as compared with CRA-2014. The similarity in total releases is primarily due to the controlling effect of Castile brine inflows on brine pressure and saturation in the waste panel and the relatively small impact that complete healing of waste panel access drifts, non-waste area drifts, and associated DRZs has on the effects of those inflows. As in SEN2, the only scenarios where DBRs were found likely to occur involve a Castile brine encounter and the subsequent release of Castile brine into the waste panel. 

The EPA also observes that a comparison between SEN3 and CRA-2014 results can be used to evaluate the effectiveness of the open drift modeling approach in CRA-2014 in achieving its objective of increasing brine saturation and thereby conservatively increasing repository releases. The comparison of results in Figure 23 shows that the open drift modeling approach in CRA-2014 does achieve its objective of increasing brine saturation compared with the closed drift approach in SEN3 in each of the four modeling scenarios illustrated in the figure. However as previously noted in Section 3.1, the greater repository volume in the open drift modeling approach also has the effect of reducing repository pressure which would tend to negate the effect of increased brine saturations on initiating DBRs. Figure 23a shows that the average pressure reduction in the open drift approach, while not substantial, reduces the likelihood of a DBR by lowering the average pressure to below the threshold for a DBR in three of the four modeling scenarios. By comparison, the SEN3 results show that in the more realistic closed drift approach the average pressure is above the threshold for a DBR in three out of the four scenarios and the likelihood of a DBR is therefore increased. Although the competing effects of the open drift approach on increasing versus decreasing the repository pressure were previously recognized, the SEN3 results provide a quantitative basis for concluding that the net effect of the open drift approach in CRA-2014 is to lower the average pressure below the threshold to initiate a DBR in all but the S2-BR scenario. The EPA concludes that the open drift modeling approach non-conservatively decreases the likelihood of a DBR and also decreases total repository releases. 

With respect to brine saturation, Figure 23b shows that although the average brine saturation is increased by the open drift modeling approach as intended and is above the mean residual saturation in three out of the four modeling scenarios, the average brine saturation is also above the mean residual saturation in the same three out of the four modeling scenarios in the closed drift modeling approach. This shows that the differences in brine saturation between the two modeling approaches are likely to have a smaller effect on DBR initiation than the differences in repository pressure because the differences in repository pressure tend to occur around the critical 8 MPa DBR threshold. 

The differences in likelihood for DBRs to occur among the scenarios cannot be translated directly to differences in total DBR releases because the scenarios have different frequencies of occurrence in WIPP PA. However, because the differences in average pressure between the two modeling approaches more significantly impact conditions for DBR initiation than the differences in average brine saturation, an increase in total DBR releases could be expected under the closed drift approach. In fact, Figure 26 demonstrates that total repository releases are slightly higher in the more realistic, closed drift approach and a comparison with Figure 22 shows that this increase is primarily due to increases in DBRs. These results also show that the modeling assumptions in the open drift approach are no longer conservative because although they do increase brine saturation, they no longer result in conservatively higher repository releases. 

6.0 SEN4 REVIEW

6.1 SEN4 Overview


SEN4 was designed to determine the cumulative effects of the following changes in parameter values and post-closure processes on repository behavior.

 Revised actinide solubility uncertainty factor distributions using the DATA0.FMl EQ 3/6 database
 Revised GLOBAL:PBRINE distribution
 Corrected BOREHOLE:TAUFAIL lower bound
 Corrected DRSPALL Version 1.22
 Corrected panel closure length
 Revised hydrogen sulfide stoichiometric coefficients 

Explanations of the purposes of the SEN4 changes are presented in the following paragraphs.

DATA0.FMl EQ3/6 Database with New Actinide Solubility Uncertainty Factor Distributions. In Completeness Comment 3-C-3 (EPA 2017b), EPA observed that the actinide solubility and aqueous speciation data in the DATA0.FM1 EQ3/6 chemistry database used in CRA-2014 was last updated using data available in 2002. Examples of relevant data developed since 2002 that need to be considered for inclusion in an updated database were provided in the Agency's comment. In response, the DOE revised the database and created DATA0.FM2 (Domski 2015; Xiong and Domski 2016). The revised database incorporated a modified hydromagnesite solubility and updated reactions for lead, borate, EDTA and citrate species. As part of EPA's completeness review the DOE had modified the actinide chemistry database -- DATA0.FM1 -- which was used in their CRA-2014 PA, to DATA0.FM2. Using this revised database, the DOE recalculated the actinide solubility uncertainty factors with this updated database. However, the EPA had not approved of this update. Consequently, the Agency specified use of the previously approved DATA0.FMl database in SEN4 to facilitate a comparison of the effect of the revised uncertainty factor distributions with the results of CRA-2014. 

Plutonium and americium isotopes represent a significant fraction of the activity in WIPP waste and the Am(III), Pu(III) and Pu(IV) oxidation states are important contributors to total repository releases as dissolved constituents in WIPP brine. Baseline estimates of the solubility of +III and +IV actinides calculated using the DATA0.FM1 EQ3/6 database have been entered into the PA parameter database. The uncertainty in those estimates is treated as a distributed uncertainty factor describing the variability about the estimate. The uncertainty factor distribution is based on the results of published solubility data from laboratory experiments relevant to WIPP conditions and describes the variability in actinide solubility by comparing published measured results to those of model predictions. EPA identified sets of relevant +III and +IV actinide solubility data for use in SEN4 that differed from the data sets used by the DOE to develop their uncertainty factor distributions for the CRA-2014. Identification of the Agency's updated data sets is described in EPA (2017e). The revised data sets resulted in changes to the uncertainty factor distributions. 

GLOBAL:PBRINE Distribution. The uncertain parameter GLOBAL:PBRINE is the probability that an exploratory borehole that intersects a WIPP waste panel also intersects a pressurized brine reservoir in the underlying Castile Formation. An EPA-mandated distribution representing the uncertainty in this parameter was used in the original 1996 PAVT baseline PA as well as the PABC-2004 and PABC-2009 recertification PAs. The DOE used a revised distribution described by Kirchner et al. (2012) in the CRA-2014 PA that EPA believes would underestimate this probability. EPA subsequently developed an updated distribution for GLOBAL:PBRINE that the Agency believes would better represent the uncertainty in this parameter and requested the DOE to use this updated distribution in SEN4. This update utilizes a cumulative distribution with a much broader range than the normal distribution utilized by the DOE. The Agency's updated distribution is documented in EPA (2017c). This parameter determines the frequency of Castile brine encounters which, as demonstrated in the SEN2 and SEN3 results, are important in establishing repository conditions favorable to initiating DBRs. 

BOREHOLE:TAUFAIL Lower Bound. The uncertain parameter BOREHOLE:TAUFAIL is the waste shear strength used to calculate cavings releases in WIPP PA. Cavings are solid waste materials sheared from the wall of boreholes that intersect the waste. The DOE revised the value of the lower bound of the uncertainty distribution for this parameter based on the 2.22 Pa mean result of a series of laboratory experiments on degraded waste surrogates conducted by SNL (Herrick et al. 2012; Herrick and Kirchner 2013). EPA believes that the lower bound should instead be based on the weakest relevant shear strength (1.60 Pa) measured in those experiments (EPA 2017d). The Agency therefore requested the DOE to use 1.60 Pa as the lowest bound in SEN4. Cavings releases are important because they combine with cuttings releases to dominate total high probability releases in WIPP PA. Lower values of waste shear strength will tend to increase total repository releases.

DRSPALL Version 1.22. The DRSPALL code is used in WIPP PA to calculate spallings releases. Spallings are solid waste materials driven up an intersecting borehole by high waste panel pressures. Version 1.22 corrects errors that are present in Version 1.21 of the code used in CRA-2014 (Kicker et al. 2016). The EPA requested the DOE to use the corrected version of this code in SEN4. This correction was expected to have a relatively minor effect on total repository releases because the spallings release volume is small relative to other release pathways.

Panel Closure Length. Although WIPP repository design includes two sets of panel closures between the operations area and Waste Panel 10, only one set was modeled in CRA-2014 (Zeitler 2015). EPA requested the DOE to correct this oversight by increasing the effective length of the modeled closure in SEN4. Longer panel closure lengths will tend to decrease the influence of the open drifts in the operations and experimental areas on WIPP performance. 

Hydrogen Sulfide Stoichiometric Coefficients. In Completeness Comments 2-C-3 and 2-C-5, and follow-up Comment 4-C-5 (EPA 2017b), EPA requested additional information supporting the CRA-2014 PA assumptions that all hydrogen sulfide created by CPR degradation will react with iron solids, and that hydrogen sulfide preferentially reacts with iron hydroxide (BRAGFLO Reaction 3) instead of metallic iron (BRAGFLO Reaction 4) (see CRA-2014 Appendix MASS, page MASS-57). These hydrogen sulfide reactions are: 

Fe(OH)2(s) + H2S(g) --> FeS(s) + H2O(l) 				(Reaction 3)

Fe(s) + H2S(g) --> FeS(s) + H2(g) 					(Reaction 4)

If metallic iron is passivated by H2S, neither Reaction 3 nor Reaction 4 will occur. Eliminating both reactions will reduce the brine volume produced by Reaction 3 and will increase gas pressures because gas is no longer consumed in Reaction 3. Eliminating Reaction 4 will not affect gas pressures because the H2S gas is converted to H2 gas on a mole by mole basis. The net effect of iron passivation is therefore to decrease brine production and increase gas production as compared with the CRA-2014 assumption that Reaction 3 is preferred. EPA requested the DOE to provide supporting data for their assumptions and to describe their effects on the water balance in the waste panels. The DOE responded by stating that project-specific data supporting these assumptions was not available and concluded that the stoichiometric coefficients concerning these reactions should be set to zero. As a result, the Agency requested the DOE to set the stoichiometric coefficients to zero in the SEN4 sensitivity study. This change eliminates any reaction of H2S with iron from WIPP PA and allows the H2S generated by waste degradation to remain as a gas. In terms of gas generation this change has the effect of converting from Reaction 3 to Reaction 4 because in Reaction 4 no brine is produced and the molar volume of repository gas remains the same as if the H2S remained unreacted. As seen in EPA's evaluation of SEN3 results in Section 5, repository releases are more sensitive to increasing pressure than to increasing brine volume and increased gas volumes will tend to increase total repository releases.

6.2 SEN4 Implementation

SEN4 implementation is described by Zeitler and Day (2016, Section 2). The PA models used for CRA-2014 were modified for SEN4 to incorporate the changes summarized above. SEN4 was run following essentially the same set of protocols used in the SEN2 and SEN3 studies and described in Sections 4 and 5. The analysis again simulated the undisturbed repository as well as the five standard scenarios used to represent inadvertent human intrusion after repository closure. The standard three replicates were also run. Parameter values that were unchanged in SEN4 were the same as the sampled and/or calculated values in the CRA-2014 PA on a vector by vector basis. This was done to facilitate comparison with the CRA-2014 PA results and determine the sensitivity of those results to the SEN4 changes in parameter values. The parameter changes in SEN2 and SEN3 representing drift creep closure were not made in SEN4, thus the results in SEN4 are based on the open drift model used in CRA-2014. The Agency's review and evaluation of these changes is presented below.

Actinide Solubility Uncertainty Distribution Implementation. The cumulative distributions for actinide solubility uncertainty factors are represented in WIPP PA by the parameters SOLMOD3:SOLVAR for the +III oxidation state and SOLMOD4:SOLVAR for the +IV oxidation state. The distributions of sampled values for these two parameters used in CRA-2014 and SEN4 are shown in Figures 28 and 29. These distributions are expressed in terms of the difference between the experimentally measured and modeled solubility on the x-axis versus the datasets analyzed, expressed in terms of cumulative probability, on the y-axis. All actinides in a given oxidation state are assigned the same solubility in WIPP PA and the same baseline solubility for the +III and +IV oxidation states are used in both CRA-2014 and SEN4. The Agency accepted the baseline solubility as expected values but was concerned that the uncertainty distributions used in CRA14 were not based on appropriately selected data sets. 

The figures show that the revised data sets provided by EPA shifted the distributions for both +III and +IV actinides. As a result of these shifts, the sampled values of SOLMOD3:SOLVAR are generally higher in SEN4 than in CRA14 while the sampled values of SOLMOD4:SOLVAR are generally lower in SEN4 than in CRA14. The solubility of the +III actinides will therefore be generally higher in SEN4 than in CRA-2014 and the solubility of the +IV actinides will be generally lower. The Agency expects this change will likely increase calculated total releases for pathways such as DBRs involving the flow of repository brine to the accessible environment because the solubility of  +III actinides is typically an order of magnitude higher than +IV actinides in WIPP brines. As a result, the +III actinide concentrations, using the revised uncertainty distributions, will be higher than in the CRA-2014 for a similar brine volume. EPA's screening criteria for solubility allows for more agreement between models and experimental data, providing increased confidence in the solubility data used for the calculations.

GLOBAL:PBRINE Distribution Implementation. The distribution of 300 GLOBAL:PBRINE sampled values in SEN4, shown in Figure 30, closely approximates the GLOBAL:PBRINE distribution developed by the Agency (EPA 2017c, Figure 5). Figure 30 shows significant differences in median and range between the Agency's SEN4 distribution and the DOE's CRA-2014 distribution. The lower tails of the two ranges are similar (about 0.05) but the upper tail of sampled values from the SEN4 distribution (about 0.55) is significantly higher than the upper tail of the CRA-2014 distribution (about 0.19). The median values of the two distributions are also different (about 0.25 in SEN4 as compared to 0.13 in CRA-2014). Increasing values of GLOBAL:PBRINE increase the probability of encountering a Castile brine reservoir. As concluded in EPA's review of SEN3, Castile brine encounters are important to total repository releases and the higher values of GLOBAL:PBRINE in SEN4 are expected to increase those releases.

BOREHOLE:TAUFAIL Lower Bound Implementation. The distribution of 300 BOREHOLE:TAUFAIL sampled values in SEN4 and CRA-2014 are shown in Figure 31. Because the distribution is uniform and has a wide range, the small decrease in the lower bound from 2.22 Pa to 1.60 Pa was not expected to have a significant effect. Figure 31 shows that the sampled values of BOREHOLE:TAUFAIL are only slightly lower in SEN4 than in CRA-2014 and the effect on total releases is expected to be small.

DRSPALL Version 1.22 Implementation. Version 1.22 corrects errors in the Forchterm calculations that are present in Version 1.21 of DRSPALL used in CRA-2014 and in previous PAs. These calculations implement the Forcheimer correction to Darcy flow to account for high gas flow rates. EPA requested the DOE to use the corrected version of this code in SEN4. Detailed descriptions of the code and the correction are presented in Kicker et al. (2016). Implementing the corrected code resulted in a noticeable increase in mean spallings releases as compared to the results of earlier, uncorrected versions of the code, but the effect on total releases was minor as shown in Figure 32. 

Panel Closure Length Implementation. The length of the northernmost panel closure in the BRAGFLO grid was increased from 30.48 m to 60.96 m in SEN4 to represent two sets of closures instead of a single set. the DOE had previously performed a PA calculation to examine the impact of doubling the length of the northernmost panel closure and found negligible changes to the pressures and saturations in the waste areas (Zeitler 2015). Although longer panel closure lengths will tend to decrease the influence of the open drifts in the operations and experimental areas on WIPP performance, the permeability of the surrounding DRZ was not changed. The cross section area of the panel closures is small compared with the cross section area of the DRZ such that flow through the panel closures is also relatively small. Reducing the permeability of the panel closures by 50% therefore may not have a significant effect on repository performance. This correction by itself is therefore expected to have a relatively minor effect on total repository releases.

Hydrogen Sulfide Stoichiometric Coefficient Implementation. Setting the stoichiometric coefficients concerning reactions of hydrogen sulfide to zero has the effect of removing the consequences of the two sulfidation reactions described in Section 6.1 from the BRAGFLO model. The DOE states that by removing these two reactions the expected impact on PA calculations is that less hydrogen sulfide gas will be consumed and less water will be produced in the waste areas. The Agency agrees that less water will be produced in the waste areas but notes that because the hydrogen sulfide gas is not consumed, the unconsumed gas will contribute to increasing pressure in the waste panels. Although increases in waste panel pressure are likely to be more important to total releases than increases in brine saturation, the dominant effect of Castile brine intrusions shown in the results of SEN3 suggest that this stoichiometric coefficient change may not have a large effect on total repository releases.

6.3 SEN4 Results

SNL's review of SEN4 results is presented in Zeitler and Day (2016, Section 4). The modifications to CRA-2014 requested by the Agency in SEN4 resulted in changes to all primary release mechanisms: cuttings and cavings, spallings, direct brine releases, and releases from the Culebra. The SEN4 results are presented as overall means of all realizations from the standard three replicates and include the upper and lower 95% confidence intervals on the means. Overall means from CRA-2014 are also presented for comparison. 

Cuttings and Cavings Releases. Cavings releases were affected by the Agency's requested reduction in the lower bound of the BOREHOLE:TAUFAIL waste shear strength distribution. However, because cavings releases are added to cuttings releases in WIPP PA, only the combined release is calculated as shown in Figure 33. Although the small reduction in the lower bound of the BOREHOLE:TAUFAIL distribution resulted in small increases in cuttings and cavings releases at lower probabilities, these increases are too small to have any meaningful effect on total mean releases.  

Spallings Releases. Spallings are solid waste materials driven up an intersecting borehole by high waste panel pressures. A principal driver for increased spallings releases in SEN4 is implementation of the corrected Forcheimer calculation in DRSPALL Version 1.22. However, spallings releases are a function of repository pressure at the time of intrusion and SEN4 changes that result in increases in pressure would also increase spallings releases. As seen in the SEN3 results, significant increases in repository pressure occur when Castile brine is encountered and the frequency of such encounters is determined by the distribution of the GLOBAL:PBRINE parameter. The increased likelihood of sampling higher values of GLOBAL:PBRINE in SEN4 shown in Figure 30 are expected to result in increased spallings releases. The increased length of the northernmost panel closure and the more significant removal of the iron sulfidation reactions in SEN4 also contribute to an increase in repository pressure. The combined effect of these changes is shown in Figure 34 where the larger volume, lower probability spallings releases increased by about half an order of magnitude. However, as discussed below, these lower probability releases are dominated by DBRs and the effect of this increase on total mean releases is minimal. 

Releases from the Culebra. Repository releases can occur through the flow of repository brine up an intersecting borehole and then laterally through the Culebra dolomite horizon and into the accessible environment at the WIPP land withdrawal boundary. The magnitude of Culebra releases is a function of actinide solubility, repository pressure, and brine saturation. The SEN4 changes that affect these parameters are the revised solubility uncertainty distributions, the increased likelihood of sampling higher values of GLOBAL:PBRINE, the increased length of the northernmost panel closures, and removal of the iron sulfidation reactions. As noted in the foregoing discussion of SEN3 results, the increased pressures are generally accompanied by decreases in brine saturations which would tend to reduce Culebra releases. The combined result of the SEN4 changes was to reduce brine flow into the Culebra but increase the concentration of +III actinides in that brine. As shown on Figure 35, this combination of SEN4 changes resulted in small increases in releases from the Culebra at low probabilities and small decreases at very low probabilities as compared with CRA-2014 results. Mean releases from the Culebra are smaller than mean spallings releases at all probabilities in both SEN4 and CRA-2014, and the effect of Culebra releases on mean total repository releases is therefore also too small to be meaningful.

Direct Brine Releases. Similar to releases from the Culebra, DBRs are also a function of actinide solubility, repository pressure, and brine saturation. DBRs are therefore influenced by the same set of SEN4 changes as Culebra releases: the revised solubility uncertainty distributions that increase the concentration of the more soluble +III actinides in repository brine, the increased likelihood of sampling higher values of GLOBAL:PBRINE which leads to more frequent Castile brine encounters, the increased length of the northernmost panel closures, and removal of the sulfidation reactions that leads to increased repository pressure. The combined effects of the SEN4 changes, shown in Figure 36, were to increase DBR volumes as well as the concentrations of +III actinides in the released brine resulting in a net increase in mean lower probability DBR releases of about half an order of magnitude.
 
Total Releases. Total releases are calculated by adding the releases from each release pathway in each realization. The overall mean is computed as the arithmetic mean of the mean releases from each replicate (Zeitler and Day 2016, p. 28). A comparison of overall means for the principal release pathways in the SEN4 analysis is shown in Figure 37. The figure illustrates the dominance of cuttings and cavings releases over total mean releases at high probabilities and the dominance of DBRs at low probabilities. 

Figure 38 shows total mean releases and associated confidence intervals for the SEN4 and CRA-2014 analyses. The confidence intervals were computed about the overall means using the Student's t-distribution and the overall mean CCDFs from each replicate (Zeitler and Day 2016, p. 28). The results show that although the overall mean increased by about half an order of magnitude in the SEN4 analysis, the mean release and its upper 95% confidence level remain well below the Agency's regulatory compliance points. 

Figure 39 shows total releases for all individual realizations from the three SEN4 replicate analyses. This figure shows that, in addition to the mean releases, no individual vectors exceeded the limits established at the Agency's 0.1 and 0.001 regulatory compliance points. 

6.4 EPA SEN4 Conclusions

The overall results show that total releases substantially increase at probabilities below 0.1 from CRA-2014 to SEN4 principally due to increased DBRs. At the upper compliance point probability of 0.1, the mean total release in SEN4 is increased by only 15% and the upper 95% confidence level is increased by only 18%. However, at the lower compliance point probability of 0.001, the mean total release is increased by 107% and the upper 95% confidence level is increased by 119% (Zeitler and Day 2016, p. 29).

The Agency attributes the observed increases primarily to 1) the revised solubility uncertainty distributions that increase the concentration of the more soluble +III actinides in repository brine, 2) the increased likelihood of sampling higher values of GLOBAL:PBRINE which leads to more frequent Castile brine encounters, and 3) removal of the iron sulfidation reactions that leads to increased repository pressure. Although the Agency considers each of these changes to be realistic alternatives to the approach taken in the CRA-2014 PA, the Agency also observes that the results of the SEN4 analysis continue to demonstrate that WIPP complies with the Agency's regulatory limits.

7.0 COMPLIANCE POINT REVIEW 

A comparison of calculated mean total releases and upper 95% confidence limits with the Agency's release limits for several past and present PAs illustrates the effects of the SEN4 changes and is presented in Table 9. The PABC-2009 PA is the Agency's currently approved performance assessment baseline. The PCS-2012 PA was a focused PA designed for the limited purpose of determining the effects of a revised panel closure system design on repository performance. The CRA-2014 PA is the PA currently under EPA evaluation for satisfying the 5-year recertification regulatory requirement and the SEN3 and SEN4 PA analyses were performed for the sensitivity studies described in this report. Graphical representations of the principal results of this compliance point review are presented in Figures 40 and 41.

The comparisons of calculated releases with EPA release limits in Table 9 show significant reductions in calculated releases in the CRA-2014 PA and in SEN3 followed by increases in SEN4. Also, the increase in SEN4 is more pronounced for the lower EPA compliance point than the upper compliance point. Figure 37 shows that calculated releases at the upper compliance point at a probability of 0.1 are primarily a function of cuttings and cavings while releases at the lower compliance point of 0.001 are primarily a function of DBRs. 

The reduction in calculated releases at the upper compliance point in the CRA-2014 as compared with previous PAs is the result of lower cavings releases due to increasing the lower bound of the BOREHOLE:TAUFAIL uncertainty range in the DOE's CRA-2014 PA from 0.05 Pa to 2.22 Pa (EPA 2017d). This increase was based on new laboratory experimental data showing that the previous lower bound of 0.05 was unrealistically low. EPA reviewed the new data and accepted the experimental results with the qualification, explained in Section 6.1, that the weakest relevant shear strength measured in those experiments of 1.60 Pa would be a more demonstrably bounding value. Decreasing the lower bound of the BOREHOLE:TAUFAIL distribution to 1.60 Pa was one of the changes made in SEN4. The Agency expected this change to have a minor effect on calculated cavings releases because the previous range of the distribution, from 2.22 to 77.0 Pa, was large and the decrease in the lower bound from 2.22 to 1.60 Pa would have only a small effect on this range. On the average, out of 300 samples of BOREHOLE:TAUFAIL taken in the three PA replicates, only 2 to 3 samples would be taken from this extended range. The decrease in calculated releases at the upper compliance point was therefore expected to be maintained in SEN4, as confirmed by the results shown in Table 9 and Figure 40.

The reduction in calculated releases at EPA's lower compliance point in CRA-2014 as compared with previous PAs is primarily the result of reduced actinide solubility, reduced sorption of actinides on colloids, and reduced releases of Castile brine. The +III actinide solubility uncertainty distribution was shifted toward lower values in the CRA-2014 PA relative to previous PAs, resulting in lower sampled +III actinide solubility that contributed to reduced calculated releases at EPA's lower compliance point. Reductions in total mobilized +III actinides in the CRA-2014 also occurred because of a lower microbial colloid proportionality constant and lower maximum concentration limit for Am(III), which again contributed to reduced releases at EPA's lower compliance point. The shift of the GLOBAL:PBRINE distribution toward lower values in the CRA-2014 reduced the frequency of Castile brine encounters which contributed to reductions in pressure and saturation-driven brine releases at the lower compliance point. In SEN4, the revised GLOBAL:PBRINE distribution, accompanied by changes that reflect Agency concerns about the actinide solubility uncertainty distribution and iron sulfidation reactions, resulted in increases in average calculated releases at the lower compliance point. These results are shown graphically in Figure 41. 

8.0 CONCLUSIONS

Each of the four sensitivity studies described in this report have yielded important conclusions. 

8.1 Conclusions from SEN1 

SNL was unable to perform SEN1 because of numerical problems within the BRAGFLO code that could not be overcome despite an extensive effort. Those problems did not allow the BRAGFLO Salado flow model to properly converge when the permeability of a modeled material was changed during a simulation and at the same time the capillary pressure component of the model was active. The Agency's SEN1 database was designed to simulate progressive permeability reductions due to creep closure of the WIPP operations and experimental area drifts while also modeling the effects of capillary forces on restricting brine and gas flow. This limitation was resolved in SEN2 by assuming that creep closure was already complete at the beginning of the simulation, thereby avoiding the need to model progressive permeability reductions. The Agency considered this resolution to be acceptable because the 200-year closure period is short compared with the 10,000-year regulatory time frame and because the permeability of healed salt is so extremely low that the effects of additional flow restrictions due to capillary forces on WIPP performance are small. 

In past PAs the DOE has resolved this modeling limitation by deactivating the capillary pressure model in the operations and experimental areas. The Agency is aware of and has accepted this modeling limitation in past PAs because capillary effects are negligible in open drifts and the DOE's resolution is consistent with the assumption that the operations and experimental area drifts do not creep close. However the Agency concludes that the capability to model capillary effects under conditions of progressive creep closure would improve WIPP performance modeling and should be incorporated in WIPP PA.      

8.2 Conclusions from SEN2

SEN2 was performed by assuming that salt creep closure and healing were already complete at the beginning of the BRAGFLO simulation instead of simulating the 200-year creep closure of the operations and experimental area drifts. As previously discussed, the Agency considered this assumption to be acceptable for the purposes of the SEN2 analysis. Creep closure and healing of the operations and experimental area drifts was found to increase pressure and decrease brine saturation throughout the waste area. The increases in pressure are primarily attributed to removal of the operations and experimental areas as storage volumes for gas. The decreases in brine saturation are primarily attributed to removal of those areas as sources of brine and to the effect of increased pressure on reducing brine seepage into the waste area. 

DBRs are important because they control low probability releases, they are a function of pressure and brine saturation, and they were therefore potentially affected by the SEN2 parameter changes. High probability releases are controlled by cuttings and cavings releases which were not affected by the SEN2 changes. The SEN2 results showed that DBRs are likely in scenarios involving Castile brine intrusions because of the significant increases in both pressure and saturation that occur due to the flow of brine into the waste panel. However, the changes in pressure and saturation due to the SEN2 parameter changes were too small to make any noticeable difference. The results also showed that DBRs are much less likely to occur in the remaining PA scenarios because they do not involve Castile brine intrusions. As a result, DBRs were essentially the same in SEN2 and in CRA-2014.

Spallings releases are a function of pressure but not of brine saturation and were increased by the SEN2 parameter changes that increased repository pressure. Although spallings releases were increased, it was not enough to have a meaningful effect on total repository releases. The remaining release pathways had negligible changes in releases.

Total mean releases, the upper 95% confidence limit on those means, and all individual vectors in the three replicates remained below EPA's WIPP release limits in SEN2.

8.3 Conclusions from SEN3

SEN3 evaluated the effects of creep closure and healing of the operations and experimental area drifts and associated DRZs as in SEN2 as well as closure and healing of the ROM salt waste panel closures and their associated DRZs. SEN3 therefore evaluated the effects of creep closure and healing of all non-waste drifts in the WIPP repository. Like SEN2, SEN3 was performed by assuming that creep closure and healing were already complete at the beginning of the BRAGFLO simulation. The Agency considered this assumption to be acceptable for purposes of SEN3. SEN3 revealed similar release trends to those of SEN2. However, creep closure and healing of the panel closures evaluated in SEN3 further increased pressure and decreased brine saturation throughout the waste area as compared to SEN2. The additional increases in pressure in SEN3 are primarily attributed to the increased isolation of the individual waste panels due to essentially complete healing of the waste panel closures, while the additional decreases in brine saturation are primarily attributed to the additional increases in pressure. 

The small increase in DBRs in SEN3 as compared with CRA-2014 is likely due to the increase in average pressure above the 8 MPa threshold for a DBR in Scenario S6-BF. That scenario involves a Castile brine intrusion and, as in the SEN2 analysis, DBRs were found to be primarily controlled by Castile brine encounters. The Agency's analysis of SEN3 results included the two rest-of-the-repository (ROR) areas and found that DBRs from those areas are highly unlikely due to low average brine saturations. This conclusion indicates that ROR conditions mapped from the BRAGFLO Salado flow model to the DBR grid are unlikely to support the occurrence of a DBR and that of the three boreholes modeled in the WIPP DBR analyses, DBRs are only likely to occur in boreholes that are randomly selected to penetrate the single modeled waste panel, which represents Panel 5 in the DBR grid. 

The DBR PA model was developed assuming concrete panel closures and an open repository where a relatively high permeability pathway existed between waste panels. This pathway allowed brine and gas to flow more freely and pressure to equilibrate more uniformly throughout the underground facility. Increased isolation of the waste panels began when the concrete panel closure design was replaced by the ROM salt design. Isolation was further increased by adopting more realistic parameter values for the ROM salt and the non-waste areas by assuming that, due to creep closure, these areas will be completely "healed" for the duration of the regulatory period. The SEN3 results have led to the following considerations and conclusions:
 Castile brine reservoirs can be hypothetically encountered more than once and in any waste panel; 
 The individual waste panels are increasingly isolated by tighter panel seals in both SEN3 and CRA-2014; and 
 The results of SEN3 demonstrate that the open drift assumption no longer results in conservatively higher repository releases, 
 The adequacy of the current assumptions governing Castile brine encounters and DBRs should therefore be further investigated.

Spallings releases were higher in SEN3 than in SEN2 but the increases were again too small to have a meaningful effect on total repository releases. Again, no meaningful changes occurred in other release pathways.

The SEN3 results showed that the original intent of the open drift assumption in WIPP PA modeling, to increase brine saturation in the waste panel, was achieved. However, the presumed effect of that assumption, to conservatively increase total repository releases above what would actually be expected, was not met. Total mean releases were, in fact, found to be slightly greater in SEN3 where the drifts were assumed to close than in CRA-2014 where the drifts were assumed to remain open. Despite these small increases in SEN3, total mean releases, the upper 95% confidence limit on those means, and all individual vectors in the three replicates remained below EPA's WIPP release limits. The Agency considers the SEN3 assumptions of salt creep closure and healing to be more realistic than the current open drift approach and consideration of incorporating those assumptions is expected to be a subject of future peer reviews of the WIPP PA conceptual models. 

8.4 Conclusions from SEN4

SEN4 evaluated the effects of the following six changes to the CRA-2014 parameter database.

Updated Actinide Solubility Uncertainty Factor Distributions with DATA0.FMl EQ3/6 Database. This update provided more realistic solubility uncertainty factor distributions for +III and +IV actinides. The update had the net result of increasing the sampled solubility of the +III actinides and decreasing the sampled solubility of the +IV actinides. These changes are likely to increase calculated total releases for pathways such as DBRs, which involve higher actinide concentrations in repository brine flowing to the surface because the solubility of +III actinides are typically an order of magnitude higher than +IV actinides.

Updated GLOBAL:PBRINE Distribution. This update included consideration of site-specific data and tended to increase sampled values of GLOBAL:PBRINE, which increases the probability of encountering a Castile brine reservoir. As concluded in EPA's review of SEN2 and SEN3, Castile brine encounters are important to total repository releases and the higher values of GLOBAL:PBRINE in SEN4 are expected to increase those releases.

Updated BOREHOLE:TAUFAIL Lower Bound. This update slightly decreased the lower bound of the distribution of BOREHOLE:TAUFAIL which is expected to slightly increase cavings releases. However, the small decrease is unlikely to have a significant effect on the release volume. 

Corrected DRSPALL Version 1.22. Version 1.22 corrects errors in the Forcheimer correction to Darcy flow that are present in Version 1.21 of DRSPALL used in CRA-2014 and in previous PAs. the DOE had previously shown that the effect of this correction on total releases was small.

Corrected Panel Closure Length. The combined lengths of the northernmost panel closures in the BRAGFLO grid was increased from 30.48 m to 60.96 m in SEN4 to represent two sets of closures instead of a single set. the DOE had previously shown that the effect of this correction on total releases was small.

Revised Hydrogen Sulfide Stoichiometric Coefficients. Setting the hydrogen sulfide stoichiometric coefficients to zero has the effect of increasing gas pressure in the waste panels. Although the magnitude of this effect cannot be individually determined from SEN4 results, the SEN3 results have demonstrated that any increases in pressure increase the likelihood of DBRs and therefore also increase total low probability releases.

The cumulative effects of these changes were provided by the results of the SEN4 analysis and are evaluated below by release pathway.

Cuttings and Cavings Releases. Cavings releases were affected by the Agency's requested reduction of the lower bound of the BOREHOLE:TAUFAIL waste shear strength distribution. However, because cavings releases are added to cuttings releases in WIPP PA, only the combined release is calculated. As expected, the small reduction in the lower bound did not have a meaningful effect on total mean releases.  
 
Spallings Releases. Spallings releases are affected in SEN4 by a combination of corrections to the Forcheimer calculation in DRSPALL Version 1.22 and by increases in repository pressure. Repository pressure was generally increased in SEN4 as a result of the updated distribution of GLOBAL:PBRINE, the increased length of the northernmost panel closure, and the updated iron sulfidation reaction stoichiometric coefficients. The combined effect of these changes was to increase spallings releases by about half an order of magnitude. Despite this significant increase, spallings releases remained low compared to DBRs and the effect of this increase in spallings on total mean releases was minimal. 

Releases from the Culebra. Releases from lateral flow through the Culebra dolomite are a function of actinide solubility, repository pressure, and brine saturation. These are affected by the revised solubility uncertainty distributions, the increased likelihood of sampling higher values of GLOBAL:PBRINE, the increased length of the northernmost panel closure, and removal of the iron sulfidation reactions. The combined effect of these changes on Culebra releases was too small to have a meaningful effect on total mean repository releases.

Direct Brine Releases. DBRs are a function of actinide solubility, repository pressure, and brine saturation. DBRs are therefore influenced by the same set of SEN4 changes as releases from the Culebra. Of these changes, the most significant are the revised solubility uncertainty distributions that increase the concentration of the more soluble +III actinides in repository brine, the increased likelihood of sampling higher values of GLOBAL:PBRINE, and changing the iron sulfidation reaction stoichiometric coefficients. The combined effects of these changes were to increase DBRs and therefore to also increase total mean low probability repository releases by about half an order of magnitude. 

The most significant effects on repository performance in the SEN4 analysis were on increasing DBRs and by extension, increasing total low probability repository releases. The Agency concludes that these increases were primarily the result of updating the solubility uncertainty distributions, updating the distribution of GLOBAL:PBRINE, and changing the iron sulfidation reaction stoichiometric coefficients. The remaining changes, updating the BOREHOLE:TAUFAIL lower bound, using the corrected DRSPALL Version 1.22, and correcting the panel closure length provided important updates and corrections to the performance calculation but had only a negligible effect on total mean releases. As in the previous sensitivity studies, the total mean releases, the upper 95% confidence limit on those means, and all individual vectors in the three replicates remained below the EPA's WIPP release limits in SEN4.


8.5 Summary of Conclusions

SEN1
 the DOE's BRAGFLO PA code is not capable of modeling capillary effects under conditions of changing permeability.
 The capability to model capillary effects under conditions of progressive creep closure would improve WIPP performance modeling and needs to be incorporated in WIPP PA.

SEN2
 Creep closure of the operations and experimental areas increases average pressures and decreases average brine saturation throughout the waste area.
 Increases in pressure and brine saturation following a Castile brine reservoir intrusion dominate the occurrence of DBRs and therefore also dominate total low probability repository releases.
 DBRs remained essentially unchanged in SEN2 because the increases in pressure and decreases in saturation due to the SEN2 parameter changes were small compared with the pressures and saturations resulting from Castile brine inflows.
 Spallings releases increased in SEN2 because of average pressure increases but remained too small to meaningfully affect total releases.
 No meaningful changes occurred in SEN2 in other release pathways.
 The mean total releases did not exceed the EPA's WIPP release limits, nor the 95% confidence level in that mean, or by any individual vectors in SEN2.

SEN3
 Trends for DBRs, spallings, cuttings, and total releases were similar to those observed in SEN2 but at higher levels.
 Creep closure and healing of all non-waste areas including waste panel access drifts effectively isolates individual waste panels. This further increases average pressures and decreases average brine saturation throughout the waste area as compared with SEN2 results.
 Castile brine encounters significantly affect repository performance. Increases in pressure and brine saturation following a Castile brine reservoir intrusion dominate the occurrence of DBRs and therefore also dominate total lower probability but higher consequence repository releases. 
 The model for Castile brine encounters was developed based on earlier assumptions of repository behavior and the continuing adequacy of those assumptions in a closed drift repository needs to be further investigated.
 Of the factors influencing DBRs, small increases in pressure are more important than small decreases in brine saturation because brine saturations are already more than sufficient in most post-disturbance scenarios to support a DBR.
 DBRs increased slightly in SEN3 as compared with CRA-2014 likely because of pressure increases in Scenario S6-BF that raised the average pressure above the 8 MPa threshold for a DBR to occur.
 Because of increased waste panel isolation, DBRs are only likely to occur in boreholes randomly selected to penetrate Panel 5 and are unlikely to occur in the two remaining boreholes considered in the DBR model.
 The DBR model was developed based on earlier assumptions of repository behavior and the continuing adequacy of those assumptions needs to be further investigated.
 Spallings releases increased in SEN3 as compared with both SEN2 and CRA-2014 because of average pressure increases but remained too small to meaningfully affect total releases.
 No meaningful changes occurred in SEN3 in other release pathways.
 The small increase in total low probability releases in the more realistic, closed drift SEN3 analysis indicates that the open drift modeling assumptions in CRA-2014 non-conservatively underestimate total repository releases despite the additional brine inflows provided by the open drift model. The open drift assumptions that result in underestimating WIPP releases need to be removed from WIPP PA.
 The mean total releases did not exceed the EPA's WIPP release limits, nor the 95% confidence level in that mean, or by any individual vectors in SEN3.

SEN4
 Correcting the BOREHOLE:TAUFAIL lower bound, using the corrected DRSPALL Version 1.22, and correcting the panel closure length had negligible effects on total mean releases.
 Revising the solubility uncertainty distributions, revising the distribution of GLOBAL:PBRINE, and revising the iron sulfidation reaction stoichiometric coefficients significantly increased DBRs as well as total mean low probability releases.  
 Castile brine reservoir encounters continue to dominate the occurrence of DBRs and therefore also dominate total low probability repository releases.
 The revisions made in the SEN4 study or alternative revisions acceptable to the EPA need to be incorporated in the CRA-2019 PA. Also, the model corrections made in the SEN4 study need to be incorporated in the CRA-2019 PA.
 The mean total releases did not exceed the EPA's WIPP release limits, nor the 95% confidence level in that mean, or by any individual vectors in SEN4.

The above sensitivity studies were intended to address a subset of EPA technical issues in a timely manner and would not require external peer reviews or petition a conceptual model evaluation. The Agency does not consider these studies as being inclusive in addressing all the technical issues identified in our CRA-2014 review. These results do demonstrate that when the aforementioned parameters are modified within the CRA-2014 PA, the releases remain well below the regulatory limits. However, in order to resolve many of the technical concerns identified in the Agency's review of the CRA-2014 PA, the EPA anticipates that the DOE address the issues identified by the EPA as soon as possible and ideally for the CRA-2019 PA.


REFERENCES

Camphouse, R.C., D.C. Kicker, T.B. Kirchner, J.J. Long, and J.J. Pasch 2011. Impact Assessment of SDI Excavation on Long-Term WIPP Performance, Revision 0. ERMS 555824. Sandia National Laboratories, Carlsbad, New Mexico. July.

Day, Brad 2016. Operations and experimental and experimental Area Sensitivity Study, Revision 0. ERMS 565918. Sandia National Laboratories, Waste Isolation Pilot Plant. Carlsbad, New Mexico. April.

Day, Brad, and Todd Zeitler 2016. Panel Closure System Sensitivity Study, Revision 0. ERMS 566725. Sandia National Laboratories, Waste Isolation Pilot Plant. Carlsbad, New Mexico. August.

the DOE (U.S. Department of Energy) 1996. Title 40 CFR Part 191 Compliance Certification
Application for the Waste Isolation Pilot Plant. the DOE-CAO 1996-2184. U.S. Department of
Energy, Waste Isolation Pilot Plant, Carlsbad Area Office. Carlsbad, NM. October.

the DOE (U.S. Department of Energy) 2014. Title 40 CFR Part 191 Compliance Recertification
Application for the Waste Isolation Pilot Plant. the DOE/WIPP-14-3503. U.S. Department of
Energy, Waste Isolation Pilot Plant, Carlsbad Area Office. Carlsbad, NM.

Domski, P.S. 2015. Memo AP-173, EQ3/6 Database Update: DATA0.FM2. Memorandum to WIPP 23 Records, October 27, 2015. ERMS 564914. Sandia National Laboratories, Carlsbad, NM. 

EPA (U.S. Environmental Protection Agency) 2013. Review of the DOE's Planned Change Request to Modify the WIPP Panel Closure System. Prepared by S. Cohen & Associates, Vienna, Virginia, for U.S. Environmental Protection Agency. Washington, DC. November.

EPA (U.S. Environmental Protection Agency) 2017a. Parameter Values for EPA Creep Closure Sensitivity Studies. Docket No. EPA-HQ-OAR-2014-0609. U.S. Environmental Protection Agency. Washington, DC. 

EPA (U.S. Environmental Protection Agency) 2017b. Evaluation of Responses to EPA Completeness Comments on the DOE 2014 Compliance Recertification Application. Docket No. EPA-HQ-OAR-2014-0609. U.S. Environmental Protection Agency. Washington, DC. 

EPA (U.S. Environmental Protection Agency) 2017c. Probability of Encountering Castile Brine beneath the WIPP Waste Panels Using the TDEM Block Method. Docket No. EPA-HQ-OAR-2014-0609. U.S. Environmental Protection Agency. Washington, DC. 

EPA (U.S. Environmental Protection Agency) 2017d. EPA Review of Proposed Modification to the Waste Shear Strength Parameter BOREHOLE:TAUFAIL. Docket No. EPA-HQ-OAR-2014-0609. U.S. Environmental Protection Agency. Washington, DC. 

EPA (U.S. Environmental Protection Agency) 2017e. Evaluation of the Compliance Recertification Actinide Source Term, Gas Generation, Backfill Efficacy, Water Balance and Culebra Dolomite Distribution Coefficient Values. Docket No. EPA-HQ-OAR-2014-0609. U.S. Environmental Protection Agency. Washington, DC. 

Herrick, C.G., M.D. Schuhen, D.M. Chapin, and D.C. Kicker 2012. Determining the
Hydrodynamic Shear Strength of Surrogate Degraded TRU Waste Materials as an Estimate for the Lower Limit of the Performance Assessment Parameter TAUFAIL, Revision 0. ERMS
558479. Sandia National Laboratories, Carlsbad, New Mexico.

Herrick, C.G. and T. Kirchner 2013. Follow-up to Questions Concerning TAUFAIL Flume Testing Raised during the November 14-15, 2012 Technical Exchange between the DOE and EPA. Memorandum to Chris Camphouse (SNL), January 24, 2013. ERMS 559081. Sandia National Laboratories, Carlsbad, New Mexico.

Kicker, Dwayne C., Courtney G. Herrick, Todd R. Zeitler, Bwalya Malama, David K. Rudeen, and Amy P. Gilkey 2016. DRSPALL: Impact of the Modification of the Numerical Spallings Model on Waste Isolation Pilot Plant Performance Assessment. SAND2016-0231. Sandia National Laboratories, Albuquerque, New Mexico. January.

Kirchner, T, T. Zeitler and R. Kirkes 2012. Evaluating the data in order to derive a value for GLOBAL:PBRINE. Memorandum to Records Center. ERMS 558724. Sandia National Laboratories, Carlsbad, New Mexico. December. 

Xiong, Y-L., and P.S. Domski. 2016. Updating the WIPP Thermodynamic Database. Analysis Report 8 for AP-173, Revision 1. ERMS 565730. Sandia National Laboratories, Carlsbad, New Mexico. 

Zeitler, T.R. 2015. Memo to Records: BRAGFLO calculations for updated northern-most
ROMPCS representation. ERMS 563875. Sandia National Laboratories, Carlsbad, New Mexico. 

Zeitler, Todd, and Brad Day 2016. CRA14 SEN4 Sensitivity Study, Revision 1. ERMS 567505. Sandia National Laboratories, Waste Isolation Pilot Plant. Carlsbad, New Mexico. December.
                                    FIGURES

Figure 1. SEN2 and CRA-2014 waste panel pressure (WAS_PRES) for a borehole intrusion at 350 years that does not intersect a Castile brine reservoir (from Day 2016, Figure 4-19).

Figure 2. SEN2 and CRA-2014 cumulative brine flow into the waste panel (BRNWASIC) for a borehole intrusion at 350 years that does not intersect a Castile brine reservoir (from Day 2016, Figure 4-55).


Figure 3. SEN2 and CRA-2014 waste panel pressure (WAS_PRES) for a borehole intrusion at 350 years that also intersects a Castile brine reservoir (from Day 2016, Figure 4-18).

Figure 4. SEN2 and CRA-2014 cumulative brine flow into the waste panel (BRNWASIC) for a borehole intrusion at 350 years that also intersects a Castile brine reservoir (from Day 2016, Figure 4-54).


Figure 5. SEN2 and CRA-2014 overall mean CCDFs for direct brine releases (from Day 2016, Figure 4-152).


Figure 6. SEN2 and CRA-2014 overall mean CCDFs for spallings releases (from Day 2016, Figure 4-150).




Figure 7. Comparison of overall means for SEN2 release components (from Day 2016, Figure 4-155).

Figure 8. SEN2 and CRA-2014 overall mean CCDFs for total normalized releases (from Day 2016, Figure 4-156).



Figure 9. Individual total normalized release vectors for the three replicates of the SEN2 study (from Day 2016, Figure 4-153).


Figure 10. SEN2, SEN3 and CRA-2014 waste panel pressure (WAS_PRES) for the undisturbed repository (from Day and Zeitler 2016, Figure 4-17).

Figure 11. SEN2, SEN3 and CRA-2014 cumulative brine inflow into the waste panel (BRNWASIC) for the undisturbed repository (from Day and Zeitler 2016, Figure 4-53).


Figure 12. SEN2, SEN3 and CRA-2014 brine saturation in the waste panel (WAS_SATB) for the undisturbed repository (from Day and Zeitler 2016, Figure 4-84).

Figure 13. SEN2, SEN3 and CRA-2014 pressure in the waste panel (WAS_PRES) for a borehole intrusion at 350 years that does not intersect a Castile brine reservoir (from Day and Zeitler 2016, Figure 4-19).


Figure 14. SEN2, SEN3 and CRA-2014 cumulative brine flow into the waste panel (BRNWASIC) for a borehole intrusion at 350 years that does not intersect a Castile brine reservoir (from Day and Zeitler 2016, Figure 4-55).


Figure 15. SEN2, SEN3 and CRA-2014 brine saturation in the waste panel (WAS_SATB) for a borehole intrusion at 350 years that does not intersect a Castile brine reservoir (from Day and Zeitler 2016, Figure 4-86).



Figure 16. SEN2, SEN3 and CRA-2014 pressure in the waste panel (WAS_PRES) for a borehole intrusion at 350 years that also intersects a Castile brine reservoir (from Day and Zeitler 2016, Figure 4-18).


Figure 17. SEN2, SEN3 and CRA-2014 cumulative brine flow into the waste panel (BRNWASIC) for a borehole intrusion at 350 years that also intersects a Castile brine reservoir (from Day and Zeitler 2016, Figure 4-54).


Figure 18. SEN2, SEN3 and CRA-2014 brine saturation in the waste panel (WAS_SATB) for a borehole intrusion at 350 years that also intersects a Castile brine reservoir (from Day and Zeitler 2016, Figure 4-85).

Figure 19. SEN2, SEN3 and CRA-2014 pressure in the waste panel (WAS_PRES) for the S6-BF scenario of a double borehole intrusion at 1,000 and 2,000 years (from Day and Zeitler 2016, Figure 4-20).


Figure 20. SEN2, SEN3 and CRA-2014 cumulative brine flow into the waste panel (BRNWASIC) for the S6-BF scenario of a double borehole intrusion at 1,000 and 2,000 years (from Day and Zeitler 2016, Figure 4-56).


 
Figure 21. SEN2, SEN3 and CRA-2014 brine saturation in the waste panel (WAS_SATB) for the S6-BF scenario of a double borehole intrusion at 1,000 and 2,000 years (from Day and Zeitler 2016, Figure 4-87).



Figure 22. SEN2, SEN3 and CRA-2014 overall mean CCDFs for direct brine releases (from Day and Zeitler 2016, Figure 4-158).
Threshold Pressure for a DBR





















Figure 23a. Comparative average pressures.

Figure 23b. Comparative average brine saturations.

Figure 23. Comparative average pressure and brine saturations in the modeled waste panel for four BRAGFLO modeling scenarios.

Figure 24. SEN2, SEN3 and CRA-2014 overall mean CCDFs for spallings releases (from Day and Zeitler 2016, Figure 4-156).

Figure 25. Comparison of overall means for SEN3 release components (from Day and Zeitler 2016, Figure 4-161).

Figure 26. SEN2, SEN3 and CRA-2014 overall mean CCDFs for total normalized releases (from Day and Zeitler 2016, Figure 4-162).

Figure 27. Individual total normalized release vectors for the three replicates of the SEN3 study (from Day and Zeitler 2016, Figure 4-159).




Figure 28. Distributions of actinide solubility uncertainty factor (SOLMOD3:SOLVAR) sampled values for the +III oxidation state from the three replicates of the SEN4 and CRA14 studies (from Zeitler and Day 2016, Figure 2-1).



Figure 29. Distributions of actinide solubility uncertainty factor (SOLMOD4:SOLVAR) sampled values for the +IV oxidation state from the three replicates of the SEN4 and CRA14 studies (from Zeitler and Day 2016, Figure 2-2).


Figure 30. Distributions of sampled values of the probability of encountering pressurized Castile brine (GLOBAL:PBRINE) from the three replicates of the SEN4 and CRA14 studies (from Zeitler and Day 2016, Figure 2-3).


Figure 31. Distributions of waste shear strength (BOREHOLE:TAUFAIL) sampled values from the three replicates of the SEN4 and CRA14 studies (from Zeitler and Day 2016, Figure 2-4).


Figure 32. Comparison of spallings and total mean releases using corrected DRSPALL Version 1.22 with earlier, uncorrected versions of DRSPALL (from Kicker et al. 2016, Figure 6-23). 


Figure 33. Overall mean cuttings and cavings releases and confidence intervals for SEN4 and CRA-2014 (from Zeitler and Day 2016, Figure 4-1).

Figure 34. Overall mean spallings releases and confidence intervals for SEN4 and CRA-2014 (from Zeitler and Day 2016, Figure 4-2).

Figure 35. Overall mean releases to the Culebra and associated confidence intervals for SEN4 and CRA-2014 (from Zeitler and Day 2016, Figure 4-3).

Figure 36. Overall mean direct brine releases and associated confidence intervals for SEN4 and CRA-2014 (from Zeitler and Day 2016, Figure 4-4).

Figure 37. Comparison of overall means for principal release pathways in the SEN4 analysis (from Zeitler and Day 2016, Figure 4-7).


Figure 38. Total mean releases and associated confidence intervals for SEN4 and CRA-2014 (from Zeitler and Day 2016, Figure 4-6).

Figure 39. Total releases for all individual realizations from the three SEN4 replicate analyses (from Zeitler and Day 2016, Figure 4-5).




Figure 40. Mean total releases as a percentage of EPA's upper (0.1 probability) release limit.



Figure 41. Mean total releases as a percentage of EPA's lower (0.001 probability) release limit.

                                    TABLES
                                       
Table 1. the DOE CRA-2014 and the EPA Proposed SEN1 Fluid Flow Parameter Values for Operations and Experimental Area Drifts 
                                       
                          Time Period & Material
                                   Porosity
                                 Permeability
                                    (m[2])
                                    CAP_MOD
                                   COMP_RCK
                                      KPT
                                    PC-MAX
                                     PCT_A
                                    PCT_EXP
                                    PO_MIN
                                   PORE_DIS
                                   RELP_MOD
                                   SAT_IBRN
                                   SAT_RBRN
                                   SAT_RGAS
                       the DOE CRA-2014 Parameter Values
                                 -5 to 0 years
                                   Cavity 3
                                     1.00
                                    1.0E-10
                                       1
                                       0
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                      11
                                       0
                                       0
                                       0
                               0 to 10,000 years
                                 OPS & EXP
                                     0.18
                                    1.0E-11
                                       1
                                       0
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                      11
                                       0
                                       0
                                       0
                      EPA Proposed SEN1 Parameter Values
                                 -5 to 0 years
                                   Cavity 3
                                     1.00
                                    1.0E-10
                                       1
                             No capillary effects
                                       0
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                      11
                                       0
                                       0
                                       0
                                 0 to 50 years
                                   OPSEX_T1
                                     0.50
                                   4.12E-13
                                       
                                       1
                             No capillary effects
                                   7.41E-10
                                  Same as DRZ
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                      11
                                       0
                                       0
                                       0
                                50 to 200 years
                                   OPSEX_T2
                                     0.18
                                   3.75E-15
                                       
                                       1
                             No capillary effects
                                   7.41E-10
                                  Same as DRZ
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                      11
                                       0
                                       0
                                       0
                              200 to 10,000 years
                                   OPSEX_T3
 CRA-2014 Sampled value for S_HALITE + (1/2) standard deviation for intact halite
      One order of magnitude higher than sampled value for intact halite
                                       2
              Activate capillary model  -  Same as intact halite
                    Same as sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                       1
                                       
                                      0.3
                        Same as mean for intact halite
                                      0.2
                        Same as mean for intact halite
Notes:
SEN1 values that differ from CRA-2014 values
Sampled values are values sampled for CRA-2104.
Intact halite has material name S_HALITE.
Table 2. the DOE CRA-2014 and SEN1 Fluid Flow Parameter Values for DRZ Adjoining Operations and Experimental Area Drifts 

                          Time Period & Material
                                   Porosity
                                 Permeability
                                    (m[2])
                                    CAP_MOD
                                   COMP_RCK
                                      KPT
                                    PC-MAX
                                     PCT_A
                                    PCT_EXP
                                    PO_MIN
                                   PORE_DIS
                                   RELP_MOD
                                   SAT_IBRN
                                   SAT_RBRN
                                   SAT_RGAS
                       the DOE CRA-2014 Parameter Values
                                 -5 to 0 years
                                     DRZ_0
                   Sampled value for intact halite + 0.0029
                                    1.0E-17
                                       1
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                       1
                                      0.0
                                      0.0
                               0 to 10,000 years
                                     DRZ_1
                   Sampled value for intact halite + 0.0029
                                    Sampled
                                       1
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                   Not Used
                                      0.0
                                      0.0
                      EPA Proposed SEN1 Parameter Values
                                 -5 to 0 years
                                     DRZ_0
                   Sampled value for intact halite + 0.0029
                                    1.0E-17
                                       1
                             No capillary effects
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                       0
                                      0.0
                                      0.0
                                0 to 500 years
                                     DRZ_1
                   Sampled value for intact halite + 0.0029
                            Sampled value for DRZ_1
                                       1
                             No capillary effects
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                       0
                                      0.0
                                      0.0
                              500 to 10,000 years
                                   DRZ_OE_2
                        Sampled value for intact halite
      One order of magnitude higher than sampled value for intact halite
                                       2
                           Activate capillary model
                             Same as intact halite
                        Sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                       1
                                       
                                      0.3
                        Same as mean for intact halite
                                      0.2
                        Same as mean for intact halite
Notes: Sampled values are values sampled for CRA-2104.; Intact halite has material name S_HALITE.
SEN1 values that differ from CRA-2014 values
Table 3. the DOE CRA-2014, SEN2 and SEN3 Fluid Flow Parameter Values for Operations and Experimental Area Drifts 
                                       
                          Time Period & Material
                                   Porosity
                                 Permeability
                                    (m[2])
                                    CAP_MOD
                                   COMP_RCK
                                      KPT
                                    PC-MAX
                                     PCT_A
                                    PCT_EXP
                                    PO_MIN
                                   PORE_DIS
                                   RELP_MOD
                                   SAT_IBRN
                                   SAT_RBRN
                                   SAT_RGAS
                       the DOE CRA-2014 Parameter Values
                                 -5 to 0 years
                                   Cavity 3
                                     1.00
                                    1.0E-10
                                       1
                                       0
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                      11
                                       0
                                       0
                                       0
                               0 to 10,000 years
                                 OPS & EXP
                                     0.18
                                    1.0E-11
                                       1
                                       0
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                      11
                                       0
                                       0
                                       0
                     Actual SEN2 and SEN3 Parameter Values
                                 -5 to 0 years
                                   Cavity 3
 CRA-2014 Sampled value for S_HALITE + (1/2) standard deviation for intact halite
      One order of magnitude higher than sampled value for intact halite
                                       2
                           Activate capillary model 
                             Same as intact halite
                    Same as sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                     0.95
                                       
                                      0.6
                       Same as maximum for intact halite
                                     0.398
                                       
                               0 to 10,000 years
                                 OPS & EXP
 CRA-2014 Sampled value for S_HALITE + (1/2) standard deviation for intact halite
      One order of magnitude higher than sampled value for intact halite
                                       2
                           Activate capillary model
                             Same as intact halite
                    Same as sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                     0.95
                                       
                                      0.6
                       Same as maximum for intact halite
                                     0.398
                                       
Notes:
SEN2 and SEN3 values that differ from CRA-2014 values
Sampled values are values sampled for CRA-2104.
Intact halite has material name S_HALITE.


                                       
                                       
Table 4. the DOE CRA-2014, SEN2 and SEN3 Fluid Flow Parameter Values for DRZ Adjoining Operations and Experimental Area Drifts 

                          Time Period & Material
                                   Porosity
                                 Permeability
                                    (m[2])
                                    CAP_MOD
                                   COMP_RCK
                                      KPT
                                    PC-MAX
                                     PCT_A
                                    PCT_EXP
                                    PO_MIN
                                   PORE_DIS
                                   RELP_MOD
                                   SAT_IBRN
                                   SAT_RBRN
                                   SAT_RGAS
                       the DOE CRA-2014 Parameter Values
                                 -5 to 0 years
                                     DRZ_0
                   Sampled value for intact halite + 0.0029
                                    1.0E-17
                                       1
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                       1
                                      0.0
                                      0.0
                               0 to 10,000 years
                                     DRZ_1
                   Sampled value for intact halite + 0.0029
                                    Sampled
                                       1
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                   Not Used
                                      0.0
                                      0.0
                     Actual SEN2 and SEN3 Parameter Values
                                 -5 to 0 years
                                   DRZ_OE_0
                        Sampled value for intact halite
                        Sampled value for intact halite
                                       2
                          Activate capillary model  
                             Same as intact halite
                        Sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                     0.95
                                       
                                      0.6
                       Same as maximum for intact halite
                                     0.398
                                       
                               0 to 10,000 years
                                   DRZ_OE_2
                        Sampled value for intact halite
                        Sampled value for intact halite
                                       2
                           Activate capillary model 
                             Same as intact halite
                        Sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                     0.95
                                       
                                      0.6
                       Same as maximum for intact halite
                                     0.398
                                       
Notes:
SEN2 and SEN3 values that differ from CRA-2014 values
Sampled values are values sampled for CRA-2104.
Intact halite has material name S_HALITE.


Table 5. the DOE CRA-2014 and the EPA Proposed SEN3 Fluid Flow Parameter Values for Waste Panel Closures the DOE
                          Time Period & Material
                                   Porosity
                                   ΦROM (%)
                                 Permeability
                                  (log m[2])
                                    CAP_MOD
                                   COMP_RCK
                                      KPT
                                    PC-MAX
                                     PCT_A
                                    PCT_EXP
                                    PO_MIN
                                   PORE_DIS
                                   RELP_MOD
                                   SAT_IBRN
                                   SAT_RBRN
                                    SROMBR
                                   SAT_RGAS
                                    SROMGR
                       the DOE CRA-2014 Parameter Values
                                 -5 to 0 years
                                   Cavity 4
                                      100
                                      -10
                                       1
                                       0
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                      11
                                       0
                                       0
                                       0
                                0 to 100 years
                                    PCS_T1
                              Sampled 6.6 to 18.7
                                    Uniform
                                   Sampled 
                                  -20.84 to 
                                    -12.00
                                    Uniform
                                       1
                                     8E-11
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                             Sampled 0.11 to 8.10
                                  Cumulative
                                       4
                                      n/a
                                   Sampled 
                                  0.0 to 0.6
                                  Cumulative
                              Sampled 0.0 to 0.4
                                    Uniform
                               100 to 200 years
                                    PCS_T2
                              Sampled 2.5 to 7.5
                                    Uniform
                                     -18.6
                                       1
                                     8E-11
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                             Sampled 0.11 to 8.10
                                  Cumulative
                                       4
                                      n/a
                                   Sampled 
                                  0.0 to 0.6
                                  Cumulative
                              Sampled 0.0 to 0.4
                                    Uniform
                              200 to 10,000 years
                                    PCS_T3
                              Sampled 0.1 to 5.19
                                    Uniform
                                     -19.1
                                       1
                                     8E-11
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                             Sampled 0.11 to 8.10
                                  Cumulative
                                       4
                                      n/a
                                   Sampled 
                                  0.0 to 0.6
                                  Cumulative
                              Sampled 0.0 to 0.4
                                    Uniform


Table 5. the DOE CRA-2014 and the EPA Proposed SEN3 Fluid Flow Parameter Values for Waste Panel Closures 
                                  (continued)
                          Time Period & Material
                                   Porosity
                                   ΦROM (%)
                                 Permeability
                                  (log m[2])
                                    CAP_MOD
                                   COMP_RCK
                                      KPT
                                    PC-MAX
                                     PCT_A
                                    PCT_EXP
                                    PO_MIN
                                   PORE_DIS
                                   RELP_MOD
                                   SAT_IBRN
                                   SAT_RBRN
                                    SROMBR
                                   SAT_RGAS
                                    SROMGR
                      EPA Proposed SEN3 Parameter Values
                                 -5 to 0 years
                                      T0
                                     0.30
                                      10
                                       2
                           Activate capillary model 
                             Same as intact halite
                                       0
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      2.1
                                       4
                             Same as intact halite
                                     0.95
                                       
                                   Sampled 
                                  0.1 to 0.2
                                    Uniform
                            0.05*(1  -  SROMBR T0)
                                       
                                 0 to 50 years
                                      T1
                             Sampled 0.075 to 0.20
                                    Uniform
                                    Sampled
                                  Triangular
                            Mode = log k(ΦT1)MODE
                          Range = +- 1.0 Assure that
                             k(ΦT1) > k(ΦT2) 
                                   See Notes
                                       2
                           Activate capillary model 
                             Same as intact halite
                    Same as sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      2.1
                                       4
                             Same as intact halite
                                      n/a
                                   Same as 
                                  value for 
                                      T0
                             Same as value for T0
                                50 to 100 years
                                      T2
                             Sampled 0.02 to 0.075
                                    Uniform
                                  Assure that
                             ΦT1>ΦT2>ΦT3
                                    Sampled
                                  Triangular
                            Mode = log k(ΦT2)MODE
                          Range = +- 1.0 Assure that
                             k(ΦT2) > k(ΦT3)
                                  See Notes 
                                       2
                           Activate capillary model 
                             Same as intact halite
                    Same as sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      1.4
                                       4
                             Same as intact halite
                                      n/a
                  Sampled Value for T1 + either 0.15 or 0.20
                             0.2*(1  -  SROMBR T2)
                                       
                              100 to 10,000 years
                                      T3
                    Same as sampled value for intact halite
                                  log k(ΦT3)
                    Same as sampled value for intact halite
                                       2
                           Activate capillary model
                             Same as intact halite
                    Same as sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                      n/a
                                       
                                  Value for 
                                   T2 + 0.25
                              0.8*(1- SROMBR T3)
                                       
                                       

SEN3 values that differ from CRA-2014 values
Notes:
For sampling permeability, use equation k(ΦTx)MODE = 1.5x10-23.5 ΦTx[8.5] with sampled value of  ΦROM  as the input variable for time period (Tx) to determine the mode of the triangular distribution, then sample a final value k(ΦTx) from the triangular distribution with a range equal to +- 1 order of magnitude from the calculated mode. Porosity ΦTx is expressed as a percentage in the equation. Equation k(ΦTx)MODE calculates permeability in m[2].
Sampled values for intact halite are values sampled for CRA-2104. Intact halite has material name S_HALITE.


                                       
Table 6. the DOE CRA-2014 and the EPA Proposed SEN3 Fluid Flow Parameter Values for DRZ Adjoining Waste Panel Closures the DOE
                          Time Period & Material
                                   Porosity
                                   ΦDRZ (%)
                                 Permeability
                                  (log m[2])
                                    CAP_MOD
                                   COMP_RCK
                                      KPT
                                    PC-MAX
                                     PCT_A
                                    PCT_EXP
                                    PO_MIN
                                   PORE_DIS
                                   RELP_MOD
                                   SAT_IBRN
                                   SAT_RBRN
                                    SDRZBR
                                   SAT_RGAS
                                    SDRZGR
                       the DOE CRA-2014 Parameter Values
                                 -5 to 0 years
                                     DRZ_0
                    Sampled value for intact halite + 0.29
                                      -17
                                       1
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                       1
                                      0.0
                                      0.0
                                0 to 200 years
                                     DRZ_1
                    Sampled value for intact halite + 0.29
                                   Sampled 
                                   -19.40 to
                                     -12.50
                                    Uniform
                                       1
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                      n/a
                                      0.0
                                      0.0
                              200 to 10,000 years
                                    DRZ_PCS
                    Sampled value for intact halite + 0.29
                                   Sampled 
                                  -20.699 to
                                     -17.00
                                  Triangular
                                       1
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                      n/a
                                      0.0
                                      0.0


Table 6. the DOE CRA-2014 and the EPA Proposed SEN3 Fluid Flow Parameter Values for DRZ Adjoining Waste Panel Closures 
                                  (continued)

                      EPA Proposed SEN3 Parameter Values
                                 -5 to 0 years
                                      T0
                                   Sampled 
                                  1.6 +- 1.0
                                    Uniform
                                    Sampled
                                 -14.22 +- 1.0
                                    Uniform
                                       2
                          Activate capillary model  
                             Same as intact halite
                        Sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                     2.27
                                       4
                             Same as intact halite
                                     0.95
                                       
                          Sampled 0.05 to 0.1 Uniform
                            0.04*(1  -  SDRZBR T0)
                                 0 TO 50 years
                                      T1
                          Same as sampled value for 
                                      T0
                               Same as sampled 
                                  value for 
                                      T0
                                       2
                          Activate capillary model  
                             Same as intact halite
                        Sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                     2.27
                                       4
                             Same as intact halite
                                      n/a
                          Same as sampled  value for 
                                      T0
                              Same as value for 
                                      T0
                                50 to 100 years
                                      T2
                           Sampled value for DRZ_PCS
                                  Assure that
                             ΦT1>ΦT2>ΦT3
                           Sampled value for DRZ_PCS
                                  Assure that
                               kT1>kT2>kT3
                                       2
                          Activate capillary model  
                             Same as intact halite
                        Sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      1.4
                                       4
                             Same as intact halite
                                      n/a
                        Sampled Value for T1 + either 
                                  0.1 or 0.2
                                     0.5*
                                   SDRZBR T2
                              100 to 10,000 years
                                      T3
                                     ΦT3
                        Sampled value for intact halite
                                      kT3
                                   Sampled 
                                  value for 
                                 intact halite
                                       2
                           Activate capillary model 
                             Same as intact halite
                        Sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                      n/a
                          Same as value for SROMBR T3
                                       
                          Same as value for SROMGR T3
                                       

Notes:
SEN3 values that differ from CRA-2014 values
Sampled values for intact halite and DRZ_PCS are values sampled for CRA-2104. Intact halite has material name S_HALITE.


Table 7. the DOE CRA-2014 and Actual SEN3 Fluid Flow Parameter Values for Waste Panel Closures 
                          Time Period & Material
                                   Porosity
                                 Permeability
                                  (log m[2])
                                    CAP_MOD
                                   COMP_RCK
                                      KPT
                                    PC-MAX
                                     PCT_A
                                    PCT_EXP
                                    PO_MIN
                                   PORE_DIS
                                   RELP_MOD
                                   SAT_IBRN
                                   SAT_RBRN
                                    SROMBR
                                   SAT_RGAS
                                    SROMGR
                       the DOE CRA-2014 Parameter Values
                                 -5 to 0 years
                                   Cavity 4
                                     1.00
                                      -10
                                       1
                                       0
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                      11
                                       0
                                       0
                                       0
                                0 to 100 years
                                    PCS_T1
                            Sampled 0.066 to 0.187
                                    Uniform
                                   Sampled 
                                  -20.84 to 
                                    -12.00
                                    Uniform
                                       1
                                     8E-11
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                             Sampled 0.11 to 8.10
                                  Cumulative
                                       4
                                      n/a
                                   Sampled 
                                  0.0 to 0.6
                                  Cumulative
                              Sampled 0.0 to 0.4
                                    Uniform
                               100 to 200 years
                                    PCS_T2
                            Sampled 0.025 to 0.075
                                    Uniform
                                     -18.6
                                       1
                                     8E-11
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                             Sampled 0.11 to 8.10
                                  Cumulative
                                       4
                                      n/a
                                   Sampled 
                                  0.0 to 0.6
                                  Cumulative
                              Sampled 0.0 to 0.4
                                    Uniform
                              200 to 10,000 years
                                    PCS_T3
                            Sampled 0.001 to 0.0519
                                    Uniform
                                     -19.1
                                       1
                                     8E-11
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                             Sampled 0.11 to 8.10
                                  Cumulative
                                       4
                                      n/a
                                   Sampled 
                                  0.0 to 0.6
                                  Cumulative
                              Sampled 0.0 to 0.4
                                    Uniform
                         Actual SEN3 Parameter Values
                                 -5 to 0 years
                                   Cavity 5
                    Same as sampled value for intact halite
                    Same as sampled value for intact halite
                                       2
                           Activate capillary model 
                             Same as intact halite
                    Same as sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                     0.95
                                       
                                   Sampled 
                                  0.5 to 0.65
                                    Uniform
                                0.8*(1- SROMBR)
                                       
                               0 to 10,000 years
                                    PCS_T1
                    Same as sampled value for intact halite
                    Same as sampled value for intact halite
                                       2
                           Activate capillary model
                             Same as intact halite
                    Same as sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                     0.95
                                       
                                   Sampled 
                                  0.5 to 0.65
                                    Uniform
                                0.8*(1- SROMBR)
                                       

                                       
Notes:
SEN3 values that differ from CRA-2014 values
Sampled values are values sampled for CRA-2104. Intact halite has material name S_HALITE.


                                       
                                       
Table 8. the DOE CRA-2014 and Actual SEN3 Fluid Flow Parameter Values for DRZs Adjoining Waste Panel Closures

                          Time Period & Material
                                   Porosity
                                 Permeability
                                  (log m[2])
                                    CAP_MOD
                                   COMP_RCK
                                      KPT
                                    PC-MAX
                                     PCT_A
                                    PCT_EXP
                                    PO_MIN
                                   PORE_DIS
                                   RELP_MOD
                                   SAT_IBRN
                                   SAT_RBRN
                                    SDRZBR
                                   SAT_RGAS
                                    SDRZGR
                       the DOE CRA-2014 Parameter Values
                                 -5 to 0 years
                                     DRZ_0
                   Sampled value for intact halite + 0.0029
                                      -17
                                       1
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                       1
                                      0.0
                                      0.0
                                0 to 200 years
                                     DRZ_1
                   Sampled value for intact halite + 0.0029
                                   Sampled 
                                   -19.40 to
                                     -12.50
                                    Uniform
                                       1
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                      n/a
                                      0.0
                                      0.0
                              200 to 10,000 years
                                    DRZ_PCS
                   Sampled value for intact halite + 0.0029
                                   Sampled 
                                  -20.699 to
                                     -17.00
                                  Triangular
                                       1
                                   7.41E-10
                                       0
                                     1E08
                                       0
                                       0
                                    101325
                                      0.7
                                       4
                                      n/a
                                      0.0
                                      0.0
                         Actual SEN3 Parameter Values
                                 -5 to 0 years
                                   DRZ_PC_0
                        Sampled value for intact halite
                        Sampled value for intact halite
                                       2
                          Activate capillary model  
                             Same as intact halite
                        Sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                     0.95
                                       
                                    SROMBR
                                    SROMGR 
                               0 to 10,000 years
                                   DRZ_PC_1
                        Sampled value for intact halite
                        Sampled value for intact halite
                                       2
                           Activate capillary model 
                             Same as intact halite
                        Sampled value for intact halite
                                       0
                             Same as intact halite
                                     1E08
                             Same as intact halite
                                     0.56
                             Same as intact halite
                         -0.346 Same as intact halite
                                    101325
                             Same as intact halite
                                      0.7
                             Same as intact halite
                                       4
                             Same as intact halite
                                     0.95
                                       
                                    SROMBR
                                       
                                    SROMGR
                                       
Notes: Sampled values are values sampled for CRA-2104. Intact halite has material name S_HALITE.

SEN3 values that differ from CRA-2014 values


Table 9. Comparison of Mean Total Normalized Releases with the EPA Release Limit

                                   Analysis
                             Releases in EPA Units
                    Mean Total as Percent of Release Limit
                                       
                         Calculated Mean Total Release
                          Upper 95% Confidence Limit
                               EPA Release Limit
                                       
EPA Upper Compliance Point at 0.1 Probability
PABC-2009 PA
                                     0.09
                                     0.10
                                      1.0
                                      9.0
PCS-2012 PA
                                     0.098
                                     0.10
                                      1.0
                                      9.8
CRA-2014 PA
                                    0.0367
                                    0.0381
                                      1.0
                                      3.7
SEN3
                                    0.0374
                                    0.0387
                                      1.0
                                      3.7
SEN4
                                    0.0423
                                    0.0449
                                      1.0
                                      4.2
EPA Lower Compliance Point at 0.001 Probability
PABC-2009 PA
                                     0.60
                                     1.77
                                      10
                                      6.0
PCS-2012 PA
                                     1.512
                                     2.81
                                      10
                                     15.1
CRA-2014 PA
                                     0.261
                                     0.308
                                      10
                                      2.6
SEN3
                                     0.299
                                     0.387
                                      10
                                      3.0
SEN4
                                     0.541
                                     0.672
                                      10
                                      5.4


