	

                                       
                                       
                                       
                        EQ3/6 COMPUTER CODE EVALUATION
                                       
                                       
                                       
                          Contract Number EP-D-10-042
                           Work Assignment No. 1-02
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                 Prepared for:
                                       
                     U.S. Environmental Protection Agency
                      Office of Radiation and Indoor Air
                        1301 Constitution Avenue, N.W.
                             Washington, DC 20460
                                       
                                Jonathan Walsh
                            Work Assignment Manager
                                       
                                       
                                       
                                 Prepared by:
                                       
                           S. Cohen & Associates
                            1608 Spring Hill Road 
                                   Suite 400
                            Vienna, Virginia 22182
JUNE 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. 

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. 
Table of Contents

1.0	Preface	i
2.0	Background	1
3.0	Role of EQ3/6 in Performance Assessment	1
4.0	DOE's Quality Assurance (QA) Procedures	2
5.0	EPA Review	3
5.1	Requirements Phase	28
5.2	Design Phase	28
5.3	Implementation Phase	29
5.4	Validation Phase	31
5.5	Installation and Checkout Phase	33
5.6	Production Software and/or Baseline Document Change Control	34
5.7	System Software and Hardware Change Control	36
6.0	EPA Findings	41
7.0	References	41
A.0	ATTACHMENT A:  DATABASE MIGRATION	1
A.1	Atomic Weights	1
A.2	Strict Basis Species	2
A.3	Auxiliary Basis Species	2
A.4	Other Aqueous Species and Solid Phases	3
A.5	Gas Phase Data	5
A.6	Pitzer Parameters	5
A.6.1	Debye-Hückel parameter A[φ]	5
A.6.2	Single Electrolyte Parameters	6
A.6.3	Common-Ion Two-Electrolyte Parameters	7
A.6.4	Neutral Ion Parameters	7
A.7	Summary of Database Review Findings	7

List of Tables

Table 1.	Summary of Test Cases	9
Table 2.	Requirements Coverage by Test Case	10
Table 3. 	Summary  of Test Cases for Unit Tests (#1 through #14) and for Verification Tests (#15 through #19) for Migration from Version 8.0 to Version 8.0a	13
Table 4.	Requirements Coverage by Test Case for Unit Tests (#1 through #14) and Verification Tests (#15 through #19) for Migration from Version 8.0 to Version 8.0a	14
Table 5.	Organic Ligand Concentrations Used in Test Cases #11 and #12, PABC09 Actinide Solubility Calculations and Test Cases #13 and #14	21
Table 6.	NP 19-1 Criteria Matrix for User's Manual	30
Table 7.	NP 19-1 Criteria Matrix for the VVP/VD	33





Acronyms and Abbreviations

Ah 		redox affinity
AMWTP	Advanced Mixed Waste Treatment Project
ANSI		American National Standards Institute
ASME 	American Society of Mechanical Engineers
CaCO3		Calcite
Ca(OH)2	Calcium hydroxide
CAG 		Compliance Application Guidance
CARD 	Compliance Application Review Document
CCA 		Compliance Certification Application
CFR		Code of Federal Regulations
Cl		chloride
CMS 		Configuration Management System
CO2(g)		Carbon Dioxide
CO3[2-]		Carbonate ion
CPR 		Cellulosics, Plastics, and Rubber
CPU 		Central Processing Unit
CRA 		Compliance Recertification Applications
DD 		Design Document
DOE 		Department of Energy
DRZ	Disturbed Rock Zone
EDTA		Ethylenediaminetetraacetic acid
Eh		Oxidation-reduction potential
EPA 		Environmental Protection Agency
FMT 		Fracture Matrix Transport
FR		Federal Register
g/L		grams per liter
GWB		Generic Weep Brine, simulated Salado brine formulation
HCl		Hydrogen chloride
HCO3[-]		Bicarbonate ion
I&C 		Installation and Checkout
ID 		Implementation Document
IEEE		Institute of Electrical and Electronics Engineers
kg		kilogram
L		liter
LLNL 		Lawrence Livermore National Laboratory
MgCl2		Magnesium chloride
Mg(OH)2	Magnesium hydroxide
MgO 		Magnesium Oxide
NaCl		Sodium chloride
NBS		National Bureau of Standards
NQA		Nuclear Quality Assurance
ODE		Ordinary Differential Equation
PA		Performance Assessment
PABC 		Performance Assessment Baseline Calculations
PAVT 		Performance Assessment Verification Test
pcH		negative logarithm of hydrogen ion concentration
pe		negative logarithm of the electron activity
pH		negative logarithm of the hydrogen ion activity
pmH		negative logarithm of the hydrogen ion molality
psi		pounds per square inch
QA 		Quality Assurance
RD 		Requirements Document
SCM		Software Configuration Management
SNL 		Sandia National Laboratories
SPR 		Software Problem Report
SQA 		Software Quality Assurance
SQAP 		Software QA Plan
TDS		Total dissolved solids
TST		Transition State Theory
UM 		User's Manual
VD 		Validation Document
VVP 		Verification and Validation Plan
WIPP 		Waste Isolation Pilot Plant


                        EQ3/6 COMPUTER CODE EVALUATION
                                       
Background

The Compliance Application Review Document (CARD) for 40 CFR  194.22 (i.e., Quality Assurance) developed during the Compliance Certification Application (CCA) discusses how the Environmental Protection Agency (EPA) verified that the development of the Department of Energy's (DOE's) computer codes for performance assessments (PA) and compliance assessments adheres to the quality assurance (QA) requirements specified in § 194.23 (i.e., Models and Computer Codes), and that the documentation developed by Sandia National Laboratories (SNL) is consistent with the requirements of the American Society of Mechanical Engineers (ASME) NQA-2a-1990 addenda, part 2.7 (ASME 1990).  

EPA has reviewed documentation provided by SNL for EQ3/6, a computer code planned to support future Compliance Recertification Applications (CRA), for consistency with the QA requirements of ASME NQA-2a-1990 addenda, part 2.7, to ASME NQA-2-1989 edition.  EPA reviewed EQ3/6 with respect to QA software, software requirements documentation, software design and implementation documentation, software verification and validation documentation, and user documentation. This review was performed in 2011 when DOE replaced the FMT code with EQ3/6 and is now included as documentation for the CRA-2014.

EQ3/6 was developed by Dr. Thomas J. Wolery at the Lawrence Livermore National Laboratory (LLNL) (Wolery 1979).  EQ3NR is the reaction path code in EQ3/6.  It calculates reaction paths (chemical evolution) in reacting water-rock and water-rock-waste systems.  Speciation in aqueous solution is an integral part of these calculations.  EQ6 computes titration processes (including fluid mixing), irreversible reaction in closed systems, irreversible reaction in some simple kinds of open systems, and heating or cooling processes, as well as solving "single-point" thermodynamic equilibrium problems.  A reaction path calculation normally involves a sequence of thermodynamic equilibrium calculations.  Chemical evolution is driven by a set of irreversible reactions (i.e., reactions out of equilibrium) and/or changes in temperature and/or pressure.  These irreversible reactions usually represent the dissolution or precipitation of minerals or other solids.  EQ3/6 computes the appearance and disappearance of phases in solubility equilibrium with the water.  Supporting software includes a data file preprocessor EQPT, conversion programs XCON3 and XCON6, and several thermodynamic data files.

Role of EQ3/6 in Performance Assessment 

In the prior CRAs, the computer code Fracture Matrix Transport (FMT) was used to model chemical equilibrium as discussed in Appendix PA, Attachment SOTERM.  DOE has replaced FMT with EQ3/6 for CRA-2014 after comparison tests for calculating the chemical conditions for the actinide speciation and solubility with a known inventory.  For the CRA-2009 PA calculation, however, the DOE used the same FMT results as those obtained during the 2004 Performance Assessment Baseline Calculations (PABC).

To date, the only calculations conducted with EQ3/6 and formally presented to the EPA are those performed by Brush et al. (2006).  In this work, EQ3/6 was used to model the microbial degradation of cellulosics, plastics, and rubber (CPR) in the presence of sulfate minerals, the generation of carbon dioxide [CO2(g)], and the effects of calcite (CaCO3) precipitation on net CO2(g) production.

Cellulosics, plastics, and rubber (CPR) materials are part of the Waste Isolation Pilot Plant (WIPP) waste inventory and are used as waste packaging materials and for waste emplacement.  These CPR materials could be microbially degraded during the 10,000-year WIPP regulatory period, producing CO2 and other gases.  Elevated CO2 concentrations in the repository could increase actinide solubilities by reducing brine pH and through formation of aqueous actinide carbonate complexes.  Anhydrous, granular, bulk magnesium oxide (MgO), therefore, has been included as an engineered barrier in WIPP. The MgO backfill fulfills the repository design requirement for an engineered barrier "to prevent or substantially delay the movement of water or radionuclides toward the accessible environment."  

EQ3/6 will also be used to model reactions of the MgO backfill with CO2. The MgO backfill is expected to react with CO2, thereby buffering pH and decreasing CO2 concentrations, so that actinide solubilities are constrained.  The MgO backfill is also expected to reduce gas pressures in the post-closure repository by reacting with gaseous CO2.

DOE's Quality Assurance (QA) Procedures

ASME's "Quality Assurance Requirements of Computer Software for Nuclear Facility Applications" (ASME 1990) provides the basis for EPA's review to verify if an appropriate QA program has been established and executed.  Part 2.7 provides requirements for the development, procurement, maintenance, and use of computer software, as applied to the design, construction, operation, modification, repair, and maintenance of nuclear facilities.  

Once nuclear quality assurance (NQA) part 2.7 was identified as the Software Quality Assurance (SQA) standard for WIPP project participants, SNL began implementation.  The standard was evaluated for applicability to WIPP software, and the existing SQA process was compared to the applicable sections of the standard (SNL 2004).

Part 2.7 endorses a systematic, life-cycle approach to software development and, although it does not require the use of a particular life cycle, it uses the ANSI/IEEE 1012 model (IEEE 1986) to illustrate its major components.  To ensure compliance with disposal regulations (in accordance with 40 CFR 194, § 194.22), SNL developed Software Requirements (NP 19-1) covering four primary development phases; (1) requirements phase, (2) design phase, (3) implementation phase, and (4) software verification and validation.  The objective of each of these phases is discussed below.

The requirements phase consists of defining and documenting both the functional requirements that the software must meet, and the verification and validation activities that must be performed in order to demonstrate that the computational requirements for the software are met.  Two documents are produced during this phase; the Requirements Document (RD) and the Verification and Validation Plan (VVP).  The RD contains the functional requirements that the proposed software must satisfy.  Specific requirements relate to the aspects of the system that must be simulated with a particular computer code. 
The design phase consists of developing and documenting the overall structure of the software and the reduction of the overall software structure into descriptions of how the code works.  During this phase, the software structural design may necessitate modifying the RD and VVP.  The Design Document (DD) provides the theoretical model, the mathematical model, and the major components of the software.  Because many of the PA computer codes were already developed before the PA calculations, the DD was not needed, as is the case for EQ3/6.  SNL uses the RD to document the purpose of the computer code(s) by listing the functional requirements.  SNL uses the VVP to develop a series of tests needed to show that the computer code properly performs the functional requirements listed in the RD.

The implementation phase consists of developing a source code using a programming language (e.g., FORTRAN) or other form suitable for compilation or translation into executable computer software.  The design, as described in the DD, is used as the basis for the software development, and it may need to be modified to reflect changes identified in the implementation phase.  Two documents are produced during this phase; the Implementation Document (ID) and the User's Manual (UM).  The ID provides the source code listing and describes the process performed to generate executable software, and the UM provides information that assists the user in understanding and using the code.

The verification and validation phase involves executing the functional test cases identified in the VVP to demonstrate that the developed software meets the requirements defined in the VVP.  The tests demonstrate the capability of the software to produce valid results for problems encompassing the range of permitted usage as defined by the UM.  One document, the Validation Document (VD), is produced during this phase.  The VD incorporates the test case input and output files, and evaluates the results versus the acceptance criteria in the VVP.  

DOE uses these procedures and documents to show that the PA computer codes calculate numerical models properly, and that the computer codes are free of coding errors and produce stable results.

EPA Review

ASME's "Quality Assurance Requirements of Computer Software for Nuclear Facility Applications" (ASME 1990) provides the basis for EPA's review to verify if an appropriate QA program has been established and executed by DOE.  Part 2.7 provides requirements for the development, procurement, maintenance, and use of computer software, as applied to the design, construction, operation, modification, repair, and maintenance of nuclear facilities.  To facilitate DOE's application process, EPA developed a Compliance Application Guidance (CAG) document, which is a companion to the final rule published in 61 FR 5224, February 9, 1996, "Criteria for the Certification and Re-Certification of the Waste Isolation Pilot Plant's (WIPP) Compliance with the 40 CFR Part 191 Disposal Regulations" (codified in 40 CFR Part 194). 

The CAG summarizes and explains EPA's expectations as to the format and content of compliance applications, based on the February 9, 1996, final rule (40 CFR Part 194).  The technical and legal requirements pertaining to the CCA are addressed by 40 CFR Parts 191 and 194.  The DOE's development and documentation of EQ3/6 are evaluated against the requirements of the CAG below.  

      a)	Any compliance application shall include:
      
            a.1) 	A description of conceptual models and scenario construction used to support any compliance application.

The chemical conditions conceptual model has changed since the CCA as new data have become available (SC&A 2008).  The original chemical conditions conceptual model was described in CCA Appendix SOTERM (DOE 1996).  A fundamental precept of the representation of the repository environment in the PA is that the actinide source term can be described by chemical equilibrium processes.

EPA reviewed the CCA conceptual model, and the only significant change was to the reaction expected to buffer CO2 fugacity for the purpose of calculating actinide solubilities.  Because of the potentially slow rate of conversion of hydromagnesite to magnesite, EPA required the assumption that the brucite-hydromagnesite reaction would buffer CO2 fugacities at higher levels than the brucite-magnesite buffer for the Performance Assessment Verification Test (PAVT).

DOE (2004) made some changes to the chemical conditions conceptual model for the CRA-2004:

       A minor change was made to the simulated brine Salado composition, from Brine A to generic weep brine (GWB), because GWB was believed to better represent the composition of intergranular brines seeping into the repository.  The effects of this change in brine composition on actinide solubilities were reviewed and found to be small (EPA 2006).
       Because of the availability of thermodynamic data for the organic ligands, the interaction of the actinides and ligands were modeled using FMT.  DOE stated that the solubilities of the +III and +IV actinides would not be significantly affected by acetate, citrate, oxalate, or ethylenediaminetetraacetic acid (EDTA) complexation.

EPA (2006) reviewed the CRA-2004 information, and on the basis of this review and additional data developed since the PAVT, required the following changes to the chemical conditions conceptual model for the subsequent PABC-2004:

       Microbial degradation experiments carried out for the WIPP program indicated that microbial degradation of cellulose was more likely because of the presence of microbes in WIPP that were capable of degrading cellulose.  These data also indicated that the long-term rates of CPR degradation were likely to be significantly lower than the rates used in the CCA PA, the PAVT, and the CRA-2004 PA.  Therefore, the probability of cellulose degradation was increased to 100%, and lower long-term degradation rates were used to model gas generation.
       The results of geochemical modeling indicated that the solubilities of the +III and +V actinides were increased by EDTA and oxalate complexation, respectively (EPA 2006).
       Because of information developed during the review of the AMWTP (TEA 2004), EPA believed that sulfate dissolved in brine and the dissolution of Disturbed Rock Zone (DRZ) minerals such as anhydrite and gypsum could cause sulfate reduction to be the dominant CPR degradation reaction.  As a result, the occurrence of significant amounts of methanogenesis should not be assumed for the purposes of calculating the required amounts of MgO in the repository.

No subsequent changes have been made to the chemical conditions conceptual model, so the conceptual model assumptions used in the PABC-2004 were also used for the PABC-2009.  In their EQ3/6 geochemical modeling of the CPR degradation reaction by sulfate reduction and interaction of CO2 with the MgO backfill, Brush et al. (2006) added polyhalite to the assemblage of Salado minerals used in the calculations.  In some of these calculations, Brush et al. (2006) found that the pH of the solutions was much higher than previously included in the chemical conditions conceptual model, which assumed that pH values would lie between approximately 8 and 9.  The higher pH values appeared to be caused by the inclusion of polyhalite in the Salado mineral assemblage (and limiting the availability of anhydrite from the DRZ).  These higher pH values correspond to aqueous total carbonate concentrations several orders of magnitude higher than previously calculated for the PABC-2004.  If these higher pH and total carbonate concentrations are confirmed, these results would represent a change in the conceptual model.  In particular, these results would change the way in which the engineered barrier is expected to control actinide solubilities by maintaining pH and CO2 fugacities, and by extension, total carbonate and actinide concentrations.  In the original geochemical conceptual model developed at the time of the CCA, it was assumed that reactions of Ca(OH)2 with CO2 would be overwhelmed by the reactions involving Mg(OH)2 and MgCl2, and prevent high pH values controlled by the Ca(OH)2/calcite buffer.  The modeling results of Brush et al. (2006), if confirmed, would contradict this part of the chemical conditions conceptual model.

Although the chemical conditions conceptual model evolved since the CCA because of the availability of additional information about processes, such as microbial degradation and complexation of actinides by organic ligands, changes to the conceptual model from the time of the CCA to the PABC-2009 have been relatively minor (SC&A 2008).

            a.2)	A description of plausible alternative conceptual model(s) seriously considered but not used to support such application, and an explanation of the reason(s) why such model(s) was not deemed to accurately portray performance of the disposal system.

A discussion of alternative approaches for calculating actinide concentrations is provided in CCA Appendix SOTERM (pages 21 - 22), as well as the CRA-2004 Appendix PA, Attachment SOTERM (pages 8 and 9), but the discussions are brief and lacking in details.  Additional information is found in supplementary references, including Bynum 1996, Novak et al. 1995 and 1996, and Brush et al. 2006.

A number of geochemical models of equilibrium chemistry are available for calculating actinide concentrations from solubility data.  In the CRA-2004, the computer code FMT was used to model chemical equilibrium, as discussed in Appendix PA, Attachment SOTERM.  Both FMT and EQ3/6 use the Pitzer activity coefficient formalism.  The Pitzer approach is the most accurate method for calculating activity coefficients for ionic species under conditions of high ionic strength as found in the WIPP brines. 

Other geochemical models are also available (e.g., WATEQ4F), but are generally applicable to low ionic strength solutions and cannot be applied to the Salado and Castile brines.  The implementation of alternative activity coefficient models, such as Harned's Rule and Specific Ion Interaction Theory, are discussed in Appendix SOTERM of the CCA, (pages 22 - 23).  However, the Pitzer approach has been found to be more accurate than the other methods and directly applicable to the brine solutions of the Salado and Castile formations.  Another geochemical model (PHREEQC) also contains algorithms based on the Pitzer approach and the cited databases for calculating activity coefficients under high ionic strength conditions applicable to the WIPP repository. 

Calculation results from EQ3/6 were compared to those of FMT in the RD and VVP and the VD developed for FMT (Docket A-93-02, II-G-3, Volume 6).  Although the comparisons are favorable, EQ3/6 is more widely used by the scientific community, and therefore DOE has decided to use EQ3/6 to support future CRA analyses.

            a.3) 	Documentation that:
                  	
                  	a.3.i) 	conceptual models and scenarios reasonably represent possible future states of the disposal system.

By definition, the thermodynamic approach used to calculate actinide concentrations under conditions of chemical equilibrium provides a depiction of the chemical state to which the disposal system should evolve over time (see EQ3/6 User's Manual and Appendix PA, Attachment SOTERM, for descriptions of the thermodynamic approach).  This assumption was reviewed and accepted by the Conceptual Model Peer Review Panel because of the long time frames (10,000 years) involved (Wilson et al. 1996). 

The basic purpose of EQ3/6 is to make two kinds of calculations pertinent to aqueous solutions and aqueous systems.  The first kind is called a speciation-solubility calculation (Jenne 1981).  This function is provided by EQ3NR code (Wolery 1992a).  The purpose of such a calculation is to describe the chemical and thermodynamic state of the solution using as input analytical data and/or theoretical assumptions, such as states of partial equilibrium with specified minerals.  Such a calculation utilizes the concepts of thermodynamic equilibrium, activity coefficients, and ion pairing and complexation, along with the corresponding mathematical descriptions and model parameters. 

The second kind of calculation is called a reaction path calculation.  This function is provided in EQ3/6 by the EQ6 code (Wolery and Daveler 1992a).  A reaction path calculation follows (or predicts) the evolution of a reacting system.  In EQ3/6, this usually refers to a system consisting of an aqueous solution and some minerals in partial equilibrium (the "reactants").  As the overall reaction progresses, the composition of the aqueous solution changes (including pH, Eh, etc.) and the solution may become saturated with new minerals.  These are generally allowed to form under the condition of partial equilibrium and are generally referred to as "product" minerals.  Other types of reaction path calculations are possible.  For example, the calculation might follow mixing with a second fluid (with no reactant minerals).  Here the mixing is treated as an irreversible "reaction."  Again, the solution composition changes and product minerals form.  Change in temperature, to simulate heating or cooling, can also be included in a reaction path calculation.  Conceptually, what a reaction progress calculation really deals with is a set of irreversible reactions (associated with the "reactants") and a set of reversible reactions (reactions in a state of partial equilibrium), including those describing ion pairing, complexing, and the formation and possible re-dissolution of "product" minerals. 

                  a.3.ii)	 mathematical models incorporate equations and boundary conditions which reasonably represent the mathematical formulation of the conceptual models,

To determine conditions of chemical equilibrium for a solid/solution system, the EQ3/6 model uses the criterion of minimization of free energy, given the constraints of the chemical composition of the fluid in question and principles of mass and charge balance and mass action defined for individual chemical reactions.  This method is entirely consistent with the definition of chemical equilibrium as being the state of lowest free energy.  The equations of mass action and mass balance are defined in the thermodynamic database of the EQ3/6 model in terms of specific reactions for aqueous speciation and solid phase solubility.  A review of the database migration is provided as Attachment A of this report.

The EQ3/6 code solves chemical equilibrium problems using the Pitzer activity coefficient formalism.  The Pitzer approach is the most accurate method for calculating activity coefficients for ionic species under conditions of high ionic strength as found in the WIPP brines.  EQ3/6 models the consequences of irreversible reactions in aqueous solutions.  It can also model fluid mixing and the consequences of changes in temperature.  This code operates both in a pure reaction progress frame and in a time frame.  In a time frame calculation, the user specifies rate laws for the progress of the irreversible reactions.  Otherwise, only relative rates are specified.   

The Agency concludes that the underlying formulations EQ3/6 are adequate to represent the current geochemical conceptual model.  If, however, the EQ3/6 modeling results reported by Brush et al. (2006) do represent an important change in the way in which the MgO engineered barrier controls geochemical conditions in the repository, additional thermodynamic information may be required to reasonably represent boundary conditions.

                  a.3.iii)	numerical models provide numerical schemes which enable the mathematical models to obtain stable solutions,

The governing equations that apply to speciation-solubility modeling consist of a large number of equations and corresponding unknowns.  The unknowns include the concentrations of all the species appearing in the model and their thermodynamic activity coefficients.  The corresponding equations are algebraic, and these must be solved using appropriate methods.  In EQ3NR, the set of unknowns is first reduced to a relatively small set of unknowns, from which the remaining unknowns can be calculated, as described in the UM (Wolery 1992b).  These are the primary iteration variables.  They are defined in this code as the log concentrations of the species in the active basis set.  The algebraic equations are solved by a combination of two iterative methods, which are applied in sequence.  The first method, called pre-Newton-Raphson optimization, has the characteristic of rapid convergence far from the solution to the equation(s), and slow (limiting first order) convergence near the solution.  It is used primarily to get all of the primary iteration variables within an order of magnitude of the solution.  The second method, a hybrid Newton-Raphson Method, has the characteristic of poor convergence behavior far from the solution, and very fast (limiting second order) convergence near the solution.  These methods thus complement one another. 

Although there do not appear to be any assessments that explicitly test the stability of the Newton-Raphson numerical schemes, a number of tests implicitly demonstrate that stable solutions are obtained.  For example, Test 2 in the VVP/VD (SNL 2006a) is a seawater test case using Pitzer equations.  The results of this test indicate that the saturation indices for EQ3NR are within the same range as those predicted by independent codes EQUIL, GEOCHEM, and SOLMNEQ, thereby indicating that the solver is providing stable solutions for those functions investigated during the test.

                  a.3.iv)	computer models accurately implement the numerical models, i.e., computer codes are free of coding errors and produce stable solutions,

EQ3/6 Version 8.0a is a modification of EQ3/6 Version 8.0.  EQ3/6 Versions 7.2A, 7.2B, and 7.2C are precursors to 8.0.  All of these codes have been qualified by DOE for use at WIPP for non-actinide problems, although EPA has not formally reviewed DOE's qualification of those earlier versions of the code.  

The test set for EQ3/6 Versions 7.2A, B and C consists of 15 test cases.  All of the test cases are sample problems described in Wolery (1992a) and Wolery and Daveler (1992b).  All the sample input files and output files are included in Wolery and Daveler (1992c). 

The input file format was extensively modified for EQ3/6 Version 8.0.  Therefore, the input files included in the EQ3/6 Version 8.0 software package are substituted for the input files in Wolery (1992a) and Wolery and Daveler (1992b).  The input files for Test #13 are converted from the input files included with EQ3/6 Version 7.2C, as explained in Appendix E (SNL 2006b).  The thermodynamic data files included in the EQ3/6 Version 8.0 software package are substituted for the thermodynamic data files used in the sample problems in Wolery (1992a) and Wolery and Daveler (1992a).  The output file format in Version 8.0 has also changed from Version 7.2C.  

The 15 original test cases were run with EQ3/6 Version 8.0.  Three new test cases were added to EQ3/6 Version 8.0 to test the utility codes EQPT, XCON3, and XCON6.  Table 1 provides a summary of the test cases.  Tests #1 through #8 evaluate the EQ3NR module of the code, Tests #9 through #15 evaluate the EQ6 module, and Tests #16 through #18 evaluate the utility codes.
Table 1.	Summary of Test Cases 
                                     Test
                                     Code
                                     File
                                   Database
                                  Description
                                       1
                                     EQ3NR
                                     swmaj
                                     .cmp
Sea water test case, with major cations and anions only
                                       2
                                     EQ3NR
                                    swmajp
                                     .hmw
Sea water test case, using Pitzer's equations
                                       3
                                     EQ3NR
                                   oxcalhem
                                     .cmp
Using mineral solubility constraints
                                       4
                                     EQ3NR
                                    custbuf
                                     .cmp
Calculating the composition of a custom pH buffer
                                       5
                                     EQ3NR
                                   fo2mineq
                                     .cmp
Computing oxygen fugacity from mineral equilibria
                                       6
                                     EQ3NR
                                    acidmwb
                                     .cmp
Computing eh from a redox couple
                                       7
                                     EQ3NR
                                    deadsea
                                     .hmw
Dead sea brine test case
                                       8
                                     EQ3NR
                                    swphel
                                     .hmw
Using pHCl as an input
                                       9
                                      EQ6
                                    pptmins
                                     .cmp
Finding precipitates from multiply-saturated sea water
                                      10
                                      EQ6
                                    heatqf
                                     .cmp
Calculating high temperature pH from quench pH
                                      11
                                      EQ6
                                     micro
                                     .cmp
Microcline dissolution in pH 4 HCl
                                      12
                                      EQ6
                                    microft
                                     .cmp
Microcline dissolution in a fluid-centered flow-through open system
                                      13
                                      EQ6
                                    gypnacl
                                     .hmw
Gypsum solubility in NaCl solutions
                                      14
                                      EQ6
                                    rwtitr
                                     .cmp
Alkalinity titration
                                      15
                                      EQ6
                                    pptqtz
                                     .cmp
Kinetics of quartz precipitation
                                      16
                                     EQPT
                                       
                                     .cmp
EQPT generation of binary thermodynamic data file
                                      17
                                     XCON3
                                     swmaj
                                     .cmp
XCON3 conversion of EQ3NR V7.2C input file to V8.0
                                      18
                                     XCON6
                                    pptmins
                                     .cmp
XCON6 conversion of EQ6 V7.2C input file to V8.0


Tests #1 through #15 were run with previous versions of EQ3/6.  SNL notes, however, that changes to the input and output files for these tests make standard "regression testing" difficult, so SNL has validated these tests on a case-by-case basis.  The EQ3/6 Version 8.0 output is verified by comparing model predicted output values to experimental results or to values calculated by independent codes.  These comparisons are discussed in SNL (2007).  Each of the test cases was developed to test particular functions of the computer code.  

New tests #16 through #18 are designed to validate the conversion utilities.  In each case, the file to be converted is used in another test case in the VVP\VD (SNL 2006b).  The file is converted, then the converted file is input to EQ3/6 Version 8.0, and the results are compared to the results from the original test case.  The individual test cases are discussed in Sections 6.1 through 6.8 of the VVP\VD.  The discussion identifies the thermodynamic data file and the input file used and the files generated by the test.  All output files are generated, but SNL did not examine all of the files.  Since the generated files are lengthy, only those portions of the files that are needed to evaluate the acceptance criteria are listed in the VVP\VD.  The thermodynamic data files and the test input files are included with the EQ3/6 Version 8.0 VVP\VD. 

All comparison calculations and plots presented in the VVP\VD were performed with Microsoft Excel(R).  The spreadsheets are contained in a single file, EQ36-v80-test.xls, stored in library EQ3/6 in the Configuration Management System (CMS).  The relative difference calculation used for comparing values as a percent is defined as follows: 

                                       

Where V8.0 is the value from EQ3/6 Version 8.0, and V7.0 is the value from EQ3/6 Version 7.0 (or other value being compared).  All test inputs and outputs are stored in library EQ3/6 in the CMS after the testing is complete.  Table 2 presents the relationship between the requirements and the test cases. 

Table 2.	Requirements Coverage by Test Case
                                  Requirement
                                Type and Number
                                  Test Number
                                       
                                       1
                                       2
                                       3
                                       4
                                       5
                                       6
                                       7
                                       8
                                       9
                                      10
                                      11
                                      12
                                      13
                                      14
                                      15
                                      16
                                      17
                                      18
Functional 	R.1
                                       X
                                       X
                                       
                                       
                                       
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
Functional 	R.2
                                       
                                       
                                       
                                       X
                                       X
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
Functional 	R.3
                                       
                                       
                                       X
                                       
                                       X
                                       
                                       
                                       
                                       X
                                       
                                       X
                                       X
                                       X
                                       
                                       
                                       
                                       
                                       
Functional 	R.4
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
Functional 	R.5
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       X
                                       
                                       X
                                       X
                                       
                                       
                                       
                                       
Functional 	R.6
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       
                                       
                                       
Functional 	R.7
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       
                                       
                                       
                                       
                                       
                                       
Functional 	R.8
                                       
                                       X
                                       
                                       
                                       
                                       
                                       X
                                       X
                                       
                                       
                                       
                                       
                                       X
                                       
                                       
                                       
                                       
                                       
External Int.	R.9
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
External Int.	R.10
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
External Int.	R.11
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
External Int.	R.12
                                       X
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
External Int.	R.13
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
External Int.	R.14
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       X
                                       
                                       
                                       
                                       
                                       X
                                       
                                       
                                       
External Int.	R.15
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
                                       
External Int.	R.16
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       
                                       
                                       
                                       
                                       
Additional 	A.1
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       
                                       
Additional 	A.2
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       
Additional 	A.3
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       

Testing for EQ3/6 Version 8.0 was completed and the VVP/VD indicates that the acceptance criteria for all test cases have been successfully met.  SNL also concludes that the testing verifies that EQ3/6 Version 8.0 satisfies all the requirements listed in the EQ3/6 RD (SNL 2006c) and the additional functionality listed in the VVP/VD (SNL 2006b). 

In 2009, EQ3/6 Version 8 was updated to EQ3/6 Version 8.0a (SNL 2009).  Revisions to the code were performed to: 

 Modify the software to recalculate the original Cphi, psi, and zeta parameters and use these to evaluate the corresponding terms in the activity coefficients, thus avoiding the "non-explicit psi error" described in Software Problem Report (SPR).

 Patch in the WIPP brine density model, to avoid otherwise necessary post-processing calculations using a spreadsheet.  Have the code output the density (g/L), TDS (g/L), and total concentrations of solutions (as molarities) for easy direct comparison with FMT output.

 Reincorporate the Pitzer (1975, eq. 47) approximation for higher-order electrostatic terms [involving the J(x) and J′(x) functions] back into EQ3/6 as an optional alternative to the more recent and accurate Harvie (1981, Appendix B) approximation.  Currently EQ3/6 (v. 8.0) only uses the latter approximation, which has become a de facto standard in the usage of Pitzer's equations.  Fracture Matrix Transport (FMT), however, uses only the older approximation.  The purpose of the reincorporation is to allow comparisons of EQ3/6 and FMT results using the Pitzer (1975) approximation as part of work under SNL Analysis Plan 140.  To assist in validating the reincorporation and subsequent analysis of EQ3/6-FMT comparison, add a code option to generate tables of the J(x) and J′(x) functions for both formulations.  In the future, only the Harvie (1981) approximation should be used in normal applications.

 Fix the following minor bugs known from SPRs:
   
      a.	YMP SPR001520060309.  Erroneous values are reported for the "NBS pH" when activity coefficients of aqueous species are scaled for consistency with a pH scale other than the NBS scale.  Although there is a work-around for this bug, it could affect future WIPP usage.  The fix is simple.  A pair of test case files already exist.
      
      b.	YMP SPR001420060309.  A TST (Transition State Theory) rate law can be misread from the EQ6 input file, depending on the order in which the terms are given.  There is also a work-around for this (re-order the terms), but this could affect future WIPP usage.  The fix is simple.  A pair of test case files already exist.
   
 A known problem with the database preprocessor.  The number of distinct Pitzer alpha coefficients is determined and is then to be written on the data1 file to be used as a dimensioning parameter by EQ3NR or EQ6.  The error is that the default value of two is written, regardless of the actual value.  In EQ3NR or EQ6, this leads to a memory access violation when the actual required dimension is greater than two.  A work-around for this uses a hex editor, such as XVI32.exe, to patch a sufficiently large value into the data1 file.  The data0.fmt data file has more than two distinct sets of Pitzer alpha coefficients.  The EQ3/6 software will be modified to write the correct value, so the work-around will no longer be required.  This fix is relatively simple.

 A straightforward option for inputting pmH will be added.  It is expected that pmH will be the usual pH input in most future WIPP usages of EQ3/6.  In Version 8.0, the "pH" input is treated as pmH if activity coefficient scaling to the Mesmer scale is specified via the iopg(2) input option.  The direct "pmH" input will be independent of the iopg(2) option.  This will be a relatively simple improvement to make and may help to avoid future mistakes when running the software.

 Minor changes required by a new compiler (Lahey/Fujitsu Fortran 95 5.70d).  The original compiler (Lahey Fortran 90 4.50h) is no longer available.  The new compiler is actually a different compiler (Fujitsu), not a more recent version of the old one.

Nineteen test cases were developed to test EQ3/6 Version 8.0a.  These are summarized in Table 3.  All of the problems have some degree of WIPP relevance.  Three of the test cases (swmajm, deadseaw, and gypnaclx) are modified EQ3/6 test problems.  The others are taken from previous FMT runs, and include examples of both historical test cases and actual applications.  Some, but not all, of the members of this set include actinides (Np, Am, and Th).

For purposes of code comparison, we define three types of test cases:

 Type 1:  The initial solution is pure water.  It is by definition charge-balanced.
 Type 2:  The initial aqueous solution composition is defined in a manner that guarantees charge balance, or the composition is pre-adjusted for charge balance, so that no subsequent adjustment is necessary in the code runs for which outputs will be compared.  This may be because the composition is simple (e.g., 4.0 m NaCl) or because of a previous adjustment made using one of the codes.
 Type 3:  The initial aqueous solution composition is not charge balanced.  A potential discrepancy between the codes may result from how this is dealt with.

The Type 1 examples include the test cases gypnaclx, f24vc7b3, f24vc7m, f24vc7k4, and f24vc7x.  The initial solution in each case is pure water, which is then reacted with a set of minerals.  Thus one likely cause of discrepancy (different means of addressing charge imbalance) is absent.  We note that one would expect ~1 x 10[-7] moles each of H[+] and OH[-] for 1 kg of "pure" H2O.  This is small enough that it will not matter whether or not these species are included in the elemental mole totals input to FMT.  The Type 2 test cases include swmajm and deadseaw.  In each of these, the initial aqueous solution composition has been adjusted for charge balance using a preliminary calculation (here using EQ3NR).  The modified composition [the chloride (Cl) was adjusted in these examples] then defines the actual test problem input to both codes.  Again, there should be no charge balance adjustment (or a negligible one) when the modified problem is run using either code.  The Type 3 test cases include all of the remaining test cases.  Each involves a starting aqueous solution that is not charge balanced (to which minerals may or may not be added).  This type of problem may show differences in code results due to the different means of addressing the charge imbalance.

In addition, in order to test the code migration from Version 8.0 to Version 8.0a, the following test cases from Version 8.0 are tested against Version 8.0a:  Test Case #15, taken from Test 3 of Version 8.0; Test Case #16, taken from Test 4 of Version 8.0; Test Case #17, taken from Test 9 of Version 8.0; Test Case #18, taken from Test 12 of Version 8.0; and Test Case #19, taken from Test 15 of Version 8.0.  The functional requirements covered by these test cases are listed in Table 3.  Functional requirement R.8 is covered by Test Cases 1 - 3 of Version 8.0a.  Version 8.0 Test Cases #2, #7, and #13 have been incorporated into the test suite for Version 8.0a.  Version 8.0 Test Cases #5, #6, #8, and #10 through #18 are not tested against Version 8.0a, as the functionalities for these test cases are already covered by the EQ3/6-to-FMT comparison test cases.  The present test cases replace the test cases defined for Version 8.0 of EQ3/6.


Table 3. 	Summary  of Test Cases for Unit Tests (#1 through #14) and for Verification Tests (#15 through #19) for Migration from Version 8.0 to Version 8.0a
                                     Test
                                     Code
                                  EQ3/6 file
                                   FMT File
                                  Description
                                       1
                                     EQ3NR
                                    swmajm
                                swmajm_08-27-09
Sea water test case, major cations and anions with Br and B
                                       2
                                     EQ3NR
                                   deadseaw
                               deadsea_08-27-09
Dead sea brine test case with Br
                                       3
                                      EQ6
                                   gypnaclx
                               gypnacl_01-14-09
Soubility of gypsum in a saturated NaCl solution
                                       4
                                      EQ6
                                    f24vc1
                                   fmt_test1
Speciation in WIPP SPC (Salado Primary Constituent) brine
                                      5A
                                     EQ3NR
                                   f24vc3sl
                                   fmt_test3
ThO2 (am) solubility in NaCl solutions up to 6 m at pmH 3.8
                                      5B
                                     EQ3NR
                                   f24vc3s2
                                   fmt_test3
ThO2 (am) solubility in NaCl solutions up to 6 m at pmH 5.5
                                       6
                                      EQ6
                                    f24vc7m
                                  fmt_test7a
Invariant point of aphthitate/glaserite-picromerite/schoenite-halite-sylvite in Na-K-Mg-Cl-SO4 system
                                       7
                                      EQ6
                                   f24vc7b3
                                  fmt_test7b
Invariant point of borax-teepleite-halite in Na-Cl-B4O7 system
                                       8
                                      EQ6
                                   f24vc7k4
                                  fmt_test7c
Invariant point of K-carbonate-K-Na-carbonate-sylvite in Na-K-Cl-CO3 system
                                       9
                                      EQ6
                                    f24vc7x
                                  fmt_test7d
Invariant point of halite-sylvite in Na-K-Cl system
                                      10
                                      EQ6
                                    f24vc8
                                   fmt_test8
Speciation of Am(III), Th(IV), and Np(V) in WIPP SPC brine
                                      11
                                      EQ6
                                    c4pgwb
                          fmt_cralbc_gwb_hmg_orgs_007
Solubility of Am(III), Th(IV), and Np(V) in WIPP GWB brine
                                      12
                                      EQ6
                                    c4per6
                          fmt_cralbc_er6_hmg_orgs_011
Solubility of Am(III), Th(IV), and Np(V) in WIPP ERDA-6 brine
                                      13
                                      EQ6
                                    c4pgwbx
                          fmt_edta_gwb_hmg_orgs_x_007
Solubility of Am(III), Th(IV), and Np(V) in WIPP GWB brine, assuming that the inventory of EDTA increases by a factor of 10 in comparison with the 2004 PABC inventory
                                      14
                                      EQ6
                                    c4per6x
                          fmt_edta_er6_hmg_orgs_x_011
Solubility of Am(III), Th(IV), and Np(V) in WIPP ERDA-6 brine, assuming that the inventory of EDTA increases by a factor of 10 in comparison with the 2004 PABC inventory
                                      15
                                     EQ3NR
                                   oxcalhem
                                      N/A
Using mineral solubility constraints
                                      16
                                     EQ3NR
                                    custbuf
                                      N/A
Calculating the composition of a custom pH buffer
                                      17
                                      EQ6
                                    pptmins
                                      N/A
Finding precipitates from multiply-saturated sea water
                                      18
                                      EQ6
                                    microft
                                      N/A
Microcline dissolution in a fluid-centered flow-through open system
                                      19
                                      EQ6
                                    pptqtz
                                      N/A
Kinetics of quartz precipitation
 

All comparison calculations were performed with Microsoft Excel.  There is at least one comparison spreadsheet per test case.  In instances in which variations were introduced in the EQ3/6 calculations, such as using the Pitzer (1975) approximation for higher-order electrostatic terms, additional spreadsheets are included.  The spreadsheets, along with all other files used in this analysis, are archived in class QA080A of library LIBEQ36 in the WIPP CMS.  The relative difference (in percent) between the EQ3/6 and FMT output values is calculated as:

                          ∆ =100 x EQ3∕6-FMTFMT

Where EQ3/6 is the value from EQ3/6 Version 8.0a, and FMT is the value from a corresponding FMT calculation.  If the reported FMT value is zero, the percent difference is not calculated and the affected values are not compared.  Generally this only happens when the previously noted FMT reporting cutoff of 1 x 10[-24] mole on the abundance of a species is triggered.  For intrinsically logarithmic quantities (pH, saturation indices), the absolute difference is used instead:

                               Δ = EQ3∕6-FMT

All of the EQ3/6 and FMT files are archived in CMS in the libraries of LIBEQ36 Class QA080A and LIBFMT, respectively.

Table 4 presents the relationship between the requirements and the test cases.

Table 4.	Requirements Coverage by Test Case for Unit Tests (#1 through #14) and Verification Tests (#15 through #19) for Migration from Version 8.0 to Version 8.0a
                                  Requirement
                                Type and Number
                                  Test Number

                                       1
                                       2
                                       3
                                       4
                                       5
                                       6
                                       7
                                       8
                                       9
                                      10
                                      11
                                      12
                                      13
                                      14
                                      15
                                      16
                                      17
                                      18
                                      19
Functional 	R.1
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       X
                                       X
                                       X
Functional	R.2
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       
                                       
                                       
Functional	R.3
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       
                                       X
                                       X
                                       
Functional	R.4
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       
                                       
Functional	R.5
                                       
                                       
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
                                       
                                       
Functional	R.6
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
Functional	R.7
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       
Functional	R.8
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
                                       
                                       
External Int. 	R.9
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
External Int.	R.10
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
External Int.	R.11
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
External Int.	R.12
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
                                       
External Int.	R.13
                                       
                                       
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       X
                                       X
                                       X
External Int.	R.14
                                       
                                       
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       X
                                       
                                       X
External Int.	R.15
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       X
                                       
External Int.	R.16
                                       
                                       
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
                                       
                                       
Functional 	R.17
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       X
                                       
                                       
                                       
                                       
                                       


Geochemical modeling of a series of test problems was carried out using both FMT Version 2.4 with database FMT_050405.CHEMDAT and EQ3/6 Version 8.0a with database data0.fmt.R0 (SNL 2011).  The results obtained using FMT and EQ3/6 were compared as part of the EQ3/6 verification and validation process.  SNL (2011) reported that preliminary modeling calculations identified two issues with the FMT code that could affect the comparisons.

The first of these issues is related to calculations of higher-order electrical interactions in brines.  Pitzer (1975) proposed a treatment of higher-order electrical interactions that includes a function called J(x) and its derivative J′(x) (SNL 2011, Appendix A).  Pitzer (1975) provided values for J(x) and J′(x) based on numerical integration, and these approximate values are used by FMT.  Harvie (1981) provided J(x) and J′(x) values based on a different approximation that is more accurate; this approximation was incorporated into the Harvie et al. (1984) model included in the WIPP geochemical model and is the approximation used by EQ3/6 Version 8.0.  Because the FMT source code uses the Pitzer (1975) approximation, the option to use either the Pitzer (1975) or Harvie (1981) approximation was added to EQ3/6 version 8.0a. 

The other issue that could affect the comparison of FMT and EQ3/6 modeling results is FMT's lack of a "front end" to facilitate problem definition.  The input data for an FMT modeling calculation consists of the elemental composition of the solution, and FMT does not allow the user to directly specify the input solution pH.  Fracture Matrix Transport (FMT) also does not allow a charge imbalance in a modeled solution, and typically the number of moles of elemental oxygen is adjusted to achieve charge balance.  This approach for charge balancing the solution results in changes in the masses of species such as bicarbonate ion (HCO3[-]), carbonate ion (CO3[2-]) and dissolved carbon dioxide (CO2(aq)).  Charge balancing using oxygen can also change the mass of H2O in the solution, modifying constituent concentrations because of the change in mass of solvent water.  As a consequence, SNL (2011) found it could be difficult to ensure that FMT and EQ3/6 were solving exactly the same problems.  To test whether the problem definition caused observed differences between the FMT and EQ3/6 results, SNL (2011) performed additional calculations for some test cases that used the FMT inguess files as the source of input data.

SNL (2011) compared the geochemical modeling results obtained using EQ3/6 and FMT for 19 test cases.  These test cases were examined to identify the calculations most similar to the WIPP PA calculations.  Two types of FMT actinide solubility calculations are used in WIPP PA.  The first type of calculation for which FMT is used is the calculation of the +III, +IV and +V actinide solubilities in WIPP brines.  Of the test cases reported by SNL (2011), Test Cases #11, #12, #13 and #14 are the most similar to the actinide solubility calculations in WIPP brines used in WIPP PA.  Test Cases #11 and #12 include FMT calculations performed to determine the solubilities of the +III, +IV and +V actinides for the CRA-2004 Performance Assessment Baseline Calculation (PABC04) in GWB and ERDA-6 brine, respectively.  Test Cases #13 and #14 are essentially the same calculations as Test Cases #11 and #12, with the concentration of EDTA in the brines increased by a factor of 10 over the concentration assumed for the PABC04 calculations. 

Fracture Matrix Transport (FMT) calculations have also been used to develop uncertainty ranges and probability distributions for the +III and +IV actinide solubilities used in PA.  For the PABC09, experimentally measured thorium(IV), neodymium(III) and americium(III) solubilities were compared to the thorium(IV) and americium(III) solubilities predicted using FMT and database FMT_050405.CHEMDAT (Xiong et al. 2009).  The differences between the measured and predicted solubilities were accumulated into histograms and cumulative distribution functions used to represent the +III and +IV actinide uncertainties.  For example, FMT calculations were performed to predict the solubility of ThO2(am) in 6 molal NaCl for comparison to measured solubilities reported by Rai et al. (1997) from pcH 3.02 to 5.16.  Test Cases #5A and #5B include FMT calculations carried out to predict ThO2(am) solubility in 6 molal NaCl.  These calculations are essentially the same as some of the calculations carried out to determine the +IV actinide solubility uncertainties for the PABC09. 

The comparisons of all test cases provided by SNL (2011) were reviewed.  Because Test Cases #5A, #5B, #11, #12, #13 and #14 are the most similar to geochemical modeling calculations carried out for WIPP PA, the results of these calculations are discussed below.  The acceptance criteria for linear quantities such as molalities, activities, activity coefficients or gas-phase fugacities are variations of 1% or less (Wolery 2008).  Base-ten logarithmic values, such as mineral saturation indices, are considered to agree within acceptable limits if the differences are less than 0.004 log units.  The pH (as pmH, pcH or Pitzer pH) values are considered to agree if they are within 0.01 unit. 

Test Cases #5A and #5B

Test Cases #5A and #5B are based on an FMT titration calculation that models the solubility of ThO2(am) as a function of pcH in 6 molal NaCl solution.  Test Case #5A compares the results under the most acidic conditions modeled (pmH 3.80) and Test Case #5B compares the results under the least acidic conditions modeled (pmH 5.50). 

Test Case 5A

SNL (2011) lists the following files for Test Case #5A:

 EQ3/6 files:  data1.fmt, f24vc3s1.3i, f24vc3s1.3o, f24vc3s1.3p
 FMT files:  FMT_050405.CHEMDAT, fmt_test3.in, fmt_test3.inguess, fmt_test3.out, titration.rhomin
 Excel file:  f24vc3s1_VVP-VD_Rev1.xls

The data1.fmt file is an unformatted file created from an ASCII-formatted data file (such as data0.fmt.R0) by running the EQPT program (Wolery and Jarek 2003).  Files f24vc3s1.3i, f24vc3s1.3o, and f24vc3s1.3p are the EQ3/6 input, output, and pickup files.  These files were examined, confirming that data0.fmt.R0 was the database file used in the EQ3/6 calculations and the input data were as stated for Test Case #5A.  The pickup file, although generated during the calculations, was not used for any subsequent calculations.

FMT_050405.CHEMDAT is the approved database file for WIPP PA.  The files fmt_test3.in, fmt_test3.inguess and titration.rhomin are FMT input files and fmt_test3.out isis the output file (Babb and Novak 1997).  In the *.in input file, the user sets the problem parameters and specifies the solution composition by providing the total elemental abundances.  In the *.inguess input file, the user can optionally specify the solution composition in terms of the aqueous solution species.  If the problem is defined using the *.in input file, the *.inguess file must exist, although it may be an empty (null) file.  For Test Case #5A, the inguess file was a null file and the input data were defined using elemental abundances in fmt_test3.in.  Test Case #5A was set up as a titration, which requires an additional input file that provides mineral densities.  This input file is titration.rhomin.

Fracture Matrix Transport (FMT) generates a primary output file -- in this case, fmt_test3.out.  Test Case #5A is the most acidic (pmH 3.80) end of a titration problem run using FMT.  The results listed in Tables 5.5-1, 5.5-2, 5.5-3, and 5.5-4 of SNL (2011) were verified by comparison with the FMT and EQ3/6 output files. The modeling results calculated using EQ3/6 agreed with the FMT results within the acceptance criteria for all parameters except for the molalities of H[+] and OH[-] and the activity coefficient for Th[4+] (Tables 5.5-1 through 5.5-4). 

For the results compared in Tables 5.5-1 through 5.5-4, the EQ3/6 calculation was performed using the Harvie (1981) approximation for J(x) and the FMT calculation was performed using the Pitzer (1975) approximation.  SNL (2011) repeated the EQ3/6 calculations using the Pitzer (1975) J(x) approximation to determine the extent to which the differences could be attributed to the use of the Harvie (1981) approximation in the EQ3/6 calculation.  The names for the input, output, and pickup files for this problem are unchanged from the earlier problem, f24vc3s1.3i, f24vc3s1.3o, and f24vc3s1.3p, respectively.  These files are differentiated from the earlier files of the same name by listing them in a different subdirectory (f24vc3_P75).  The input file f24vc3s1.3i for this calculation includes the string "USEOLDPITZER75," which is echoed in the f24vc3s1.3o output file in this subdirectory, specifying that the calculation used the Pitzer (1975) J(x) approximation. 

The results obtained using EQ3/6 and the Pitzer (1975) J(x) approximation are compared to the FMT results in Tables 5.5-5, 5.5-6, and 5.5-7.  These data were verified by comparison with the FMT and EQ3/6 output files.  The agreement between the FMT and EQ3/6 modeling results was improved by using the same J(x) approximation for both calculations, although the Δ value for the Th[4+] activity coefficient of 1.745% still exceeds the acceptance criteria.  The remaining differences between the FMT and EQ3/6 results are likely caused by the presence of additional water in the FMT calculation, which is likely caused by the way in which modeling problems are defined for FMT input (SNL 2011).

The total thorium concentration is the parameter of interest in PA calculations carried out to determine the range and probability distribution for the +IV actinide solubilities.  The greatest difference in the total thorium concentration was observed between FMT and EQ3/6 when the EQ3/6 calculations were performed using the Harvie (1981) J(x) approximation.  This difference was only 0.092% of the total thorium concentration calculated by FMT.  Such a small difference in the total thorium concentration will not significantly affect calculations of the thorium solubility uncertainty range and probability distribution.  Accordingly, the performance of EQ3/6 for Test Case #5A is acceptable.

Test Case #5B

SNL (2011) lists the following files for Test Case #5B:

 EQ3/6 files:  data 1.fmt, f24vc3s2.3i, f24vc3s2.3o, and f24vc3s2.3p
 FMT files:  FMT_050405.CHEMDAT, fmt_test3.in, fmt_test3.inguess, fmt_test3.out, and titration.rhomin
 Excel file:  f24vc3s2_VVP-VD_Rev1.xls

Files f24vc3s2.3i, f24vc3s2.3o and f24vc3s2.3p are the EQ3/6 input, output, and pickup files.  These files were examined, confirming that data0.fmt.R0 was the database file used in the EQ3/6 calculations and the input data were as stated for Test Case #5B5B.  The pickup file, although generated during the calculations, was not used for any subsequent calculations.  The FMT files are the same files used in Test Case #5A, because both solutions were modeled in the original FMT titration calculation.  As noted for Test Case #5A, fmt_test3.inguess is a null file and titration.rhomin is an input file containing mineral density data.

Test Case #5B is the least acidic (pmH 5.50) end of the titration calculation performed with FMT.  The results listed in Tables 5.6-1 through 5.6-4 were verified by comparison with the FMT and EQ3/6 output files.  The masses of solution and H2O calculated by FMT and EQ3/6 differ significantly (Table 5.6-1), because the FMT calculation was scaled to 1 L of solution, whereas the EQ3/6 calculation was scaled to 1 kg H2O.  Because Test Case #5B is a solubility calculation, this difference in scales does not affect the other results.  The EQ3/6 calculations performed using the Harvie (1981) J(x) approximation provided results that are within the specified acceptance limits, except for molality and activity coefficient of Th[4+] (Tables 5.6-2 and 5.6-3).  The total thorium concentrations calculated using the two codes differed by only 0.423% (Table 5.6-2).  The EQ3/6 calculation was repeated using the Pitzer (1975) J(x) approximation, and the agreement was improved.  SNL (2011) attributed the remaining small difference in Th[4+] molality (Δ = 1.033%) to slightly different problem definitions using FMT and EQ3/6. 

The maximum difference in total thorium concentrations between the FMT and EQ3/6 codes is 0.423% of the FMT-calculated concentration.  As noted for Test Case #5A, such a small difference in the total thorium concentration will not significantly affect calculation of the uncertainty range and probability distribution for the +IV actinide solubility.  Consequently, the performance of EQ3/6 is acceptable for Test Case #5B.

Test Cases #11 and #12

Test Cases #11 and #12 were performed to compare the results of the PABC04 FMT calculations of americium(III), thorium(IV) and neptunium(V) solubilities to EQ3/6 calculation results.  The Test Case #11 calculations use GWB brine and the Test Case #12 calculations use ERDA-6 brine.  The only differences between these calculations and the more recent PABC09 actinide solubility calculations are the assumed organic ligand concentrations.

Test Case #11

SNL (2011) lists the following files for Test Case #11:

 EQ3/6 files:  data1.fmt, c4pgwb.3i, c4pgwb.3o, c4pgwb.3p, c4pgwb.6i, c4pgwb.6o, and c4pgwb.6p
 FMT files:  FMT_050405.CHEMDAT, fmt_cra1bc_gwb_hmag_orgs_007.in, fmt_cra1bc_gwb_hmag_orgs_007.inguess and fmt_cra1bc_gwb_hmag_orgs_007.out
 Excel file:  c4pgwb_VVP-VD_Rev1.xls

The FMT and EQ3NR input files and the file Conc_density_calcs_EV2008_Rev1.xls (SNL 2011) were used to verify the data listed in Tables 5.12-1 and 5.12-2.  The FMT input file fmt_cra1bc_gwb_hmag_orgs_007.in includes an entry for 1.38  10[-15] moles of the pseudo-element Charge:EL that is not listed in Table 5.12-1.  This pseudo-element appears to have been included for charge balance purposes.  The EQ3NR input file includes an initial estimated pH of 6.17 that was not included in Table 5.12-2.

The EQ3NR input and output files were used to verify the data listed in Tables 5.12-3 through 5.12-8.  SNL (2011) states on page 90 regarding Table 5.12-7 that "magnesite (MgCO3) precipitates and is thus saturated."  The phase that actually precipitated in the modeling calculations is hydromagnesite5424 (Table 5.12-7).  

The masses of solution and H2O calculated by FMT and EQ3/6 (Table 5.12-4) differed significantly because the FMT calculation was scaled to 1 kg of solvent water, whereas the EQ3/6 calculation was scaled to 1 L of solution.  Because Test Case #11 is a solubility calculation, this difference in scales does not affect the other results.  The EQ3/6 calculations performed using the Harvie (1981) J(x) approximation resulted in calculated molalities (Table 5.12-5) and activity coefficients (Table 5.12-6) that differ from the FMT results by more than 1% for a fairly large number of aqueous species.  In addition, the mineral saturation indices for a number of solid phases differ by amounts greater than the 0.004 log unit tolerance (Table 5.12-7).  However, the total americium(III), neptunium(V) and thorium(IV) concentrations varied by fairly small percentages, with Δ values ranging from -1.35% to 0.8888% (Table 5.12-8).

The EQ3/6 calculation for Test Case #11 was repeated using the same Pitzer (1975) J(x) approximation as FMT (Tables 5.12-9 through 5.12-11).  Comparison of the FMT modeling results to the EQ3/6 results obtained using the Pitzer (1975) J(x) approximation showed that the differences between the aqueous species molalities and activity coefficients were greatly reduced compared to the results of EQ3/6 modeling with the Harvie (1981) approximation.  However, differences that exceed the acceptance criteria remained between the results of the two calculations.  SNL (2011) attributed the remaining differences to different definitions of the problem in the FMT and EQ3/6 input.  To minimize these differences, SNL (2011) repeated the EQ3/6 calculations using the initial solution data in the fmt_cra1bc_gwb_hmag_orgs_007.inguess file (Table 5.12-12).  The results obtained using these input data and the Pitzer (1975) J(x) approximation are compared to the FMT modeling results in Tables 5.12-13 through 5.12-17.  The differences between the FMT and EQ3/6 modeling results are quite small, demonstrating that the differences between the FMT and EQ3/6 modeling results can be satisfactorily explained by differences in the J(x) approximation used by the two codes and differences in the problem definition resulting from the way in which aqueous solutions are defined in an FMT input file. 

The total americium(III), thorium(IV) and neptunium(V) concentrations are parameters of interest in calculations carried out to determine the actinide solubilities for use in WIPP PA.  The greatest differences in these total actinide concentrations occur when the EQ3/6 calculations were performed using the Harvie (1981) J(x) approximation and the EQ3/6 input data were defined in the expected manner.  The relative differences ranged from -1.35% of the concentration calculated by FMT for neptunium(V) to 0.8888% of the concentration calculated by FMT for americium(III).  Such small differences in the total actinide concentration will not significantly affect the results of WIPP PA because the uncertainties associated with the actinide solubilities have ranges of several orders of magnitude.  Based on the results presented by SNL (2011), the performance of EQ3/6 is acceptable for Test Case #11.

Test Case #12

Test Case #12 is similar to Test Case #11, except the brine composition used in the calculations is ERDA-6 instead of GWB.  SNL (2011) lists the following files for Test Case #12:

 EQ3/6 files:  data1.fmt, c4per6.3i, c4per.3o, c4per6.3p, c4per6.6i, c4per.6o, and c4per6.6p
 FMT files:  FMT_050405.CHEMDAT, fmt_cra1bc_er6_hmag_orgs_11011.in, fmt_cra1bc_er6_hmag_orgs_11011.inguess and fmt_cra1bc_er6_hmag_orgs_11011.out
 Excel file:  c4per6_VVP-VD_Rev1.xls

The FMT and EQ3NR input files and the file Conc_density_calcs_EV2008_Rev1.xls (SNL 2011) were used to verify the data listed in Tables 5.13-1 and 5.13-2. The FMT input file fmt_cra1bc_er6_hmag_orgs_011.in includes an entry for 1.38  10[-15] moles of the pseudo-element Charge:EL that is not listed in Table 5.13-1.  This pseudo-element appears to have been included for charge balance purposes.  The EQ3NR input file includes an initial estimated pH of 6.17 that was not included in Table 5.13-2.

The FMT and EQ3NR input and output files were used to verify the data listed in Tables 5.12-3 through 5.12-8. A minor omission was noted in Table 5.13-8, where the percentage of NpO2Oxalate- (7.51%) was not included.

The modeling results for Test Case #12 obtained using FMT and EQ3/6 are compared in Tables 5.13-4 through 5.13-8 (SNL 2011).  The masses of solution and H2O calculated by FMT and EQ3/6 differ significantly because the FMT calculation was scaled to 1 kg of solvent water, whereas the EQ3/6 calculation was scaled to 1 L of solution.  Because Test Case #12 is a solubility calculation, this difference in scales does not affect the other results.  The EQ3/6 calculations performed using the Harvie (1981) J(x) approximation resulted in calculated molalities and activity coefficients that differ by more than the allowable 1% for a fairly large number of aqueous species (Tables 5.13-5 and 5.13-6).  The mineral saturation indices also differ by more than the acceptance criterion of 0.004 (Table 5.13-7).  However, the total americium(III), neptunium(V), and thorium(IV) concentrations varied by relatively small percentages compared to the aqueous speciation, with Δ values ranging from -1.0707% to 0.17% (Table 5.13-8).  Such small differences in the total actinide concentration will not significantly affect the results of WIPP PA, because the uncertainties associated with the actinide solubilities have a range of several orders of magnitude.  SNL (2011) repeated the EQ3/6 calculations for Test Case #12 using the Pitzer (1975) J(x) approximation and SNL (2011) found that the results of the comparison are improved.  Based on the results presented by SNL (2011), the performance of EQ3/6 is acceptable for Test Case #12.

Test Cases #13 and #14

Test Cases #13 and #14 compare the FMT and EQ3/6 calculations of americium(III), thorium(IV), and neptunium(V) solubilities in WIPP brines with the EDTA concentration increased by a factor of 10 over the PABC04 concentration.  The Test Case #13 and Test Case #14 calculations use GWB brine and ERDA-6 brine, respectively.  These calculations differ from the most recent WIPP PA actinide solubility calculations (PABC09) only because of different assumed organic ligand concentrations (Table 5).  The organic ligand that has the greatest effect on the WIPP PA actinide solubility calculations is EDTA.  The EDTA concentration used in the Test Case #13 and Test Case #14 calculations is only slightly greater than the PABC09 EDTA concentration.  Consequently, these test case calculations are reasonably representative of the current baseline actinide solubility calculations.
                                       
Table 5.	Organic Ligand Concentrations Used in Test Cases #11 and #12, PABC09 Actinide Solubility Calculations and Test Cases #13 and #14
                                 Calculations
                            Test Cases #11 and #12
                                   (moles/L)
                                    PABC09
                                   (moles/L)
                            Test Cases #13 and #14
                                   (moles/L)
                                    Acetate
                                1.06 x 10[-2]
                                1.94 x 10[-2]
                                1.06 x 10[-2]
                                    Citrate
                                8.06 x 10[-4]
                                2.38 x 10[-3]
                                8.06 x 10[-4]
                                     EDTA
                                8.14 x 10[-6]
                                6.47 x 10[-5]
                                8.14 x 10[-5]
                                    Oxalate
                                4.55 x 10[-2]
                                1.73 x 10[-2]
                                4.55 x 10[-2]

Test Case #13

SNL (2011) lists the following files for Test Case #13:

 EQ3/6 files:  data1.fmt, c4pgwbx.3i, c4pgwbx.3o, c4pgwbx.3p, c4pgwbx.6i, c4pgwbx.6o, and c4pgwbx.6p
 FMT files:  FMT_050405.CHEMDAT, fmt_edta_gwb_hmag_orgs_x_007.in, fmt_edta_gwb_hmag_orgs_x_007.inguess and fmt_edta_gwb_hmag_orgs_x_007.out
 Excel file:  c4pgwbx_VVP-VD-Rev1.xls

The FMT and EQ3NR input files and the file Conc_density_calcs_EV2008_Rev1.xls were used to verify the data listed in Tables 5.14-1 and 5.14-2.  The FMT input file fmt_edta_gwb_hmag_orgs_x_007.in includes 1.38  10[-15] moles of the pseudo-element Charge:EL that is not listed in Table 5.14-1.  This pseudo-element appears to have been included for charge balance purposes.  The EQ3NR input file includes an initial estimated pH of 6.17 that was not included in Table 5.14-2.  

Tables 5.14-33 through 5.14-8 were verified by comparison to the FMT and EQ3/6 output files.  The SNL (2011) states on page 124 regarding Table 5.14-7 that "magnesite (MgCO3) precipitates and is thus saturated."  The mineral phase that actually precipitated based on the data in Table 5.14-7 is hydromagnesite5424.  
The masses of solution and H2O calculated by FMT and EQ3/6 (Table 5.14-4) differ significantly because the FMT calculation was scaled to 1 kg of solvent water, whereas the EQ3/6 calculation was scaled to 1 L of solution.  Because Test Case #13 is a solubility calculation, this difference in scales does not affect the other results.  The EQ3/6 calculations performed using the Harvie (1981) J(x) approximation provided calculated molalities and activity coefficients results that differ by more than 1% for a fairly large number of aqueous species (Tables 5.14-5 and 5.14-6).  Calculated mineral saturation indices for a number of minerals differ by more than the acceptance limit of 0.004 (Table 5.14-7).  However, the total americium(III), neptunium(V), and thorium(IV) concentrations varied by relatively small percentages compared to the aqueous speciation results, with Δ values ranging from -1.3535% to 2.01% (Table 5.14-8). 

The EQ3/6 calculations for Test Case #13 were repeated using the same Pitzer (1975) J(x) approximation as FMT.  The differences between the FMT and EQ3/6 modeling results are smaller if the Pitzer (1975) J(x) approximation is used for both calculations.  The EQ3/6 calculations were also repeated using both the Pitzer (1975) J(x) approximation and defining the EQ3/6 input starting solution using the results from the fmt_edta_gwb_hmag_orgs_x_007.inguess file.  The results obtained using these input data and the Pitzer (1975) J(x) approximation compared to the FMT modeling results are presented in Tables 5.14-9 through 5.14-11.  The differences between the FMT and EQ3/6 modeling results are quite small.  These results demonstrate that the differences between the FMT and EQ3/6 modeling results can be satisfactorily explained by the different J(x) approximation used by the two codes and differences in the problem definition resulting from the way in which aqueous solutions are defined in an FMT input file.

The total americium(III), thorium(IV), and neptunium(V) concentrations are the parameters of interest in WIPP PA actinide solubility calculations.  The greatest differences in these total actinide concentrations are observed between FMT and EQ3/6 when the EQ3/6 calculations are performed using the Harvie (1981) J(x) approximation and the EQ3/6 input data are defined in the expected manner.  The relative differences ranged from -1.3535% of the concentration calculated by FMT for neptunium(V) to 2.01% of the concentration calculated by FMT for americium(III).  Such small differences in the total actinide concentrations will not significantly affect the results of WIPP PA because the uncertainties associated with the actinide solubilities have a range of several orders of magnitude.  Based on the results presented by SNL (2011), the performance of EQ3/6 is acceptable for Test Case #13.

Test Case #14

Test Case #14 is similar to Test Case #13, except the brine composition used in the calculations is ERDA-6 instead of GWB.  SNL (2011) lists the following files for Test Case #14:

 EQ3/6 files:  data1.fmt, c4per6x.3i, c4per6x.3o, c4per6x.3p, c4per6x.6i, c4per6x.6o, and c4per6x.6p
 FMT files:  FMT_050405.CHEMDAT, fmt_cra1bc_er6_hmag_orgs_x_007.in, fmt_cra1bc_er6_hmag_orgs_x _007.inguess and fmt_cra1bc_er6_hmag_orgs_x _007.out
 Excel file:  c4per6x.xls
The FMT file names listed by SNL (2011) are incorrect.  The correct file names for the FMT input and output files for Test Case #14 are fmt_cra1bc_er6_hmag_orgs_x_011.in, fmt_cra1bc_er6_hmag_orgs_x_011.inguess and fmt_cra1bc_er6_hmag_orgs_x_011.out. 

The FMT and EQ3NR input files and the file Conc_density_calcs_EV2008_Rev1.xls were used to verify the data listed in Tables 5.15-1 and 5.15-2.  The FMT input file fmt_edta_gwb_hmag_orgs_x_0011.in includes 1.38  10[-15] moles of the pseudo-element Charge:EL that is not listed in Table 5.15-1.  This pseudo-element appears to have been included for charge balance purposes. Table 5.15-1 contains several minor errors: 1) the amount of C is missing (0.016 moles), 2) an incorrect number of moles are listed for B (0.016 instead of 0.063) and 3) an incorrect number of moles are listed for Br (0.063 instead of 0.011). The EQ3NR input file includes an initial estimated pH of 6.17 that was not included in Table 5.14-2.  The values reported in Tables 5.14-3 through 5.14-8 were verified using the FMT and EQ3/6 output files.

The modeling results obtained for Test Case #14 using FMT and EQ3/6 are compared in Tables 5.15-4 through 5.15-8 (SNL 2011).  The masses of solution and H2O calculated by FMT and EQ3/6 differ significantly (Table 5.14-4) because the FMT calculation was scaled to 1 kg of solvent water, whereas the EQ3/6 calculation was scaled to 1 L of solution.  Because Test Case #14 is a solubility calculation, this difference in scales does not affect the other results. 

Calculated molalities and activity coefficients differ by more than 1% for a fairly large number of aqueous species (Tables 5.15-5 and 5.15-6).  The differences in the mineral saturation indices for a number of phases differ by more than the acceptance criterion of 0.004 (Table 5.15-7).  However, the total americium(III), neptunium(V), and thorium(IV) concentrations vary by relatively small percentages compared to the aqueous speciation data (Table 5.15-8), with Δ values ranging from -1.06% to 0.14%.  Such small differences in the total actinide concentration will not significantly affect the results of WIPP PA because the uncertainties associated with the actinide solubilities have a range of several orders of magnitude.  SNL (2011) repeated the EQ3/6 calculations for Test Case #14 using the Pitzer (1975) J(x) approximation and found that the results of the comparison were improved, as observed for other test cases.  Based on the results presented by SNL (2011), the performance of EQ3/6 is acceptable for Test Case #14.

Test Case #17

Test Case #17 is used by SNL (2011) to compare the results obtained using EQ6 versions 8.0 and 8.0a (Shoemaker 2011). SNL (2011) lists the following files for Test Case #17:

 data file:  data1.cmp
 input files:  pptmins.3i, pptmins.6i 
 output files:  pptmins.3o, pptmins.3p, pptmins.6o, pptmins.6p, pptmins.6tx, pptmins.csv 
 Excel file:  pptmins_V-VVP-VD_Rev1.xls
However, Shoemaker (2011) noted that the EQ3NR input, output and pickup files for this test case were actually swpar.3i, swpar.3o and swpar.3p. The EQ3NR output and pickup files were generated using version 8.0 and database version data0.v8.R6. The pickup file swpar.3p was used to create the EQ6 input file (pptmins.6i) that was used for both the EQ6 version 8.0 and version 8.0a calculations. 

In addition to the EQ6 output file pptmins.6o and pickup file pptmins.6p, the output files pptmins.6tx and pptmins.csv are provided. The pptmins.6tx file is a tab-separated output file and the pptmins.csv file is a comma-separated value file. The results provided for Test Case #17 (Tables 5.18-1 through 5.18-4) were verified by comparing the data in the tables to the EQ3/6 output files obtained using versions 8.0 and 8.0a. The results obtained using the two EQ3/6 versions were all within acceptance criteria.

Test Case #18

Test Case #18 is used by SNL (2011) to compare the results obtained using EQ6 versions 8.0 and 8.0a (Shoemaker 2011). SNL (2011) lists the following files for Test Case #18:

 data file:  data1.cmp
 input files:  microft.3i, microft.6i 
 output files:  microft.3o, microft.3p, microft.6o, microft.6p, microft.6tx, microft.csv 
 Excel file:  microft_V-VVP-VD_Rev1.xls
However, Shoemaker (2011) noted that the EQ3NR input, output and pickup files for this test case were actually ph4hcl.3i, ph4hcl.3o and ph4hcl.3p. The EQ3NR output and pickup files were generated using EQ3/6 version 8.0 and database version data0.v8.R6. The pickup file ph4hcl.3p was used to create the EQ6 input file (microft.6i) that was used for both the EQ6 version 8.0 and version 8.0a calculations. 

In addition to the EQ6 output file microft.6o and pickup file microft.6p, the output files microft.6tx and microft.csv are provided. The microft.6tx file is a tab-separated output file and the microft.csv file is a comma-separated value file. 

Test Case #19

Test Case #19 is used by SNL (2011) to compare the results obtained using EQ6 versions 8.0 and 8.0a (Shoemaker 2011). SNL (2011) lists the following files for Test Case #19:

 data file:  data1.cmp
 input files:  pptqtz.3i, pptqtz.6i 
 output files:  pptqtz.3o, pptqtz.3p, pptqtz.6o, pptqtz.6p, pptqtz.6tx, pptqtz.csv 
 Excel file:  pptqtz _V-VVP-VD_Rev1.xls
However, Shoemaker (2011) stated that the EQ3NR input, output and pickup files for this test case were actually sio2.3i, sio2.3o and sio2.3p. These EQ3NR output and pickup files were generated using EQ3/6 version 8.0 and database version data0.v8.R6. The pickup file sio2.3p was used to create the EQ6 input file (pptqtz.6i) used for both the EQ6 version 8.0 and 8.0a calculations.  

In addition to the EQ6 output file pptqtz.6o and pickup file pptqtz.6p, the output files pptqtz.6tx and pptqtz.csv are provided. The pptqtz.6tx file is a tab-separated output file and the pptqtz.csv file is a comma-separated value file. 

Summary of Test Case Review Findings

The test cases summarized by SNL (2011) adequately represent calculations performed for WIPP PA, and results of the test case calculations show that the EQ3/6 performance is acceptable.   A few minor errors in SNL (2011) were identified, as previously noted in this report. These errors are typographic in nature and do not affect the conclusions presented by SNL (2011):

 The file names for the FMT input and output files for Test Cases #11, #12, #13 and #14 should include the string "hmag" instead of "hmg" in Table 5-1 of SNL (2011).
 The mineral that precipitates from solution in Test Cases #11 through #14 is hydromagnesite5424 instead of magnesite as stated on pages 90, 114, 124, and 139 of SNL (2011).
 The percentage of NpO2Oxalate- (7.51%) was omitted from Table 5.13-8.
  The Excel spreadsheet for Test Case #11 is incorrectly given as c4pgwb.xls (SNL 2011, page 85), and the correct file name is cp4gwb_VVP-VD_Rev1.xls.
 The FMT input and output file names listed for Test Case #14 by SNL (2011, page 133) are fmt_edta_er6_hmag_orgs_x_007.in, fmt_edta_er6_hmag_orgs_x_007.inguess and fmt_edta_er6_hmag_orgs_x_007.out. However, the correct file names for Test Case #14 are fmt_edta_er6_hmag_orgs_x_011.in, fmt_edta_er6_hmag_orgs_x_011.inguess and fmt_edta_er6_hmag_orgs_x_011.out.
 The number of moles of C used in the FMT input (0.016 moles) is omitted from Table 5.15-1 and incorrect values are listed for B and Br (the correct values are 0.063 and 0.011 moles, respectively).
 The EQ3NR input, output and pickup file names provided by SNL (2011) for Test Case #17 are incorrect, and the correct file names are swpar.3i, swpar.3o and swpar.3p, respectively
 The EQ3NR input, output and pickup file names provided by SNL (2011) for Test Case #18 are incorrect, and the correct file names are ph4hcl.3i, ph4hcl.3o and ph4hcl.3p, respectively
 The EQ3NR input, output and pickup file names provided by SNL (2011) for Test Case #19 are incorrect, and the correct file names are sio2.3i, sio2.3o and sio2.3p, respectively

                  a.3.v)	conceptual models have undergone peer review according to § 194.27,

A fundamental aspect of the representation of the repository environment in the PA is that the actinide source term can be described by chemical equilibrium processes.  The EQ3/6 code assumes chemical equilibrium as applied to actinide chemistry that incorporates the Pitzer approach for calculating activity coefficients relevant to the high ionic strength conditions present in the Salado and Castile brines.  This conceptualization is summarized in Appendix SOTERM (Docket A-93-02, II-G-1, Volume XVII), and more detail regarding chemical equilibrium processes is provided in Docket A-93-02, II-G-1, Ref. #302, Ref. #477, Ref. #479; Docket A-93-02, II-G-3, Volume 6 (Novak et al. 1995); and Bynum 1996.

The total actinide source term used in the assessments consists of two component parts; the dissolved actinides (which includes actinides complexed with various ligands) and actinides adsorbed by various types of colloidal material.  The colloidal actinide component is, in turn, made up of several constituents; actinides sorbed on microbes, actinides sorbed on mineral fragments in the colloidal size range, actinides that have been condensed into clusters of polymerized actinide species ("intrinsic colloids"), and actinides that have adsorbed onto large organic molecules found in soils (commonly called humics).  

The dissolved actinide source term, colloidal actinide source term, and chemical conditions conceptual models were reviewed by the Conceptual Models Peer Review Panel.  The Panel concluded that all these models were adequate for use in PA (see Section 3 of the Technical Support Document  -  Models and Computer Codes § 194.23).  

Based on its review of DOE's description of the conceptual model(s) presented in the CRA and supporting documentation, the Agency concludes that the conceptual model has undergone peer review and the documentation of this conceptual model is adequate.  

      b) 	Computer codes used to support any compliance application shall be documented in a manner that complies with the requirements of ASME NQA-2a-1990 addenda, part 2.7, to ASME NQA-2-1989 edition. 

As noted in the sections that follow, EQ3/6 is documented as specified in NQA-2a-1990 in the RD, UM, and the VVP/VD.  Although these reports describe the operation and testing of the EQ3/6 code, they do not document the generation of the thermodynamic database used by EQ3/6 or the modifications to that database planned for the future.  This issue is important, because calculations of the solubilities of actinide solids with EQ3/6 are directly dependent on its thermodynamic database.  This database has undergone continuous revision since the CCA, and this has probably resulted in improvements, but the changes have made it difficult to reproduce past results exactly. 

      c)  	Documentation of all models and computer codes included as part of any compliance application performance assessment calculation shall be provided.

To demonstrate that computer software is in compliance with QA requirements outlined in § 194.22, SNL established a life-cycle management process for the software used to support PA.  Their qualification approach for the software is based on the life-cycle phases outlined in ASME NQA-2a-1990 addenda, part 2.7, which are as follows: 

 Planning
 Requirements
 Design
 Implementation
 Software Validation
 Installation and Checkout
 Operation and Maintenance
 Retirement

A Software QA Plan (SQAP) is produced during the planning phase for new software development.  Software under configuration control and developed within the scope of these QA requirements does not require a stand-alone SQAP.  Following the development of the SQAP, a strict sequence of performing activities is not required, provided that all specified requirements for each phase are met and the intent of the requirements are not subverted.  SQAPs may be written for an individual code or a set of codes.

For EQ3/6, an SQAP was developed for Version 8.0 that documents the adequacy of the pre-existing EQ3/6 (for earlier code versions) and related code documentation, and identifies the documentation that was necessary to transition this code into the operation and maintenance life-cycle phase for use in future analyses (SNL 2006d).  Following SNL's review of the documentation that existed for earlier versions of EQ3/6, SNL determined that the following
primary documents would need to be developed, reviewed, and maintained in accordance with NP 19-1 for EQ3/6:

 User's Manual (UM)
 Requirements Document (RD)
 Verification and Validation Plan/Validation Document (VVP/VD)

Since EQ3/6 is acquired software, a DD is not required.  Because the Version 8.0a source code was not provided to SNL with the software, SNL did not develop an ID. Note: SNL does have the source code for Versions 7.2(a,b,c). SNL has since completed the development of these additional documents (i.e., UM, RD, VVP/VD) and together they provide the foundation of the EQ3/6 documentation.  

To determine whether this EQ3/6 documentation satisfies the computer software documentation requirements as established by NQA-2a, part 2.7, the Agency conducted a review of the core documents as discussed below.

Requirements Phase

The document produced during the requirements phase is the RD, which identifies the computational requirements of the code (e.g., EQ3/6 must be able to simulate actinide geochemistry).  The RD also describes how the code will be tested to ensure that those requirements are satisfied.   

The RD provides a code overview indicating that EQ3/6 is a software package utilized to perform geochemical modeling encompassing fluid-mineral interactions and/or solution-mineral equilibria in aqueous systems.  The software package is composed of two major components:  EQ3NR, a speciation-solubility code, and EQ6, a reaction path modeling code to simulate water/rock interaction or fluid mixing in either a pure reaction progress mode or a time-dependent or kinetic mode.  Supporting software includes EQPT, a data file preprocessor, along with several supporting thermodynamic data files.  
The RD clearly outlines eight functional and eight external interface requirements that are tested in the VVP/VD developed for EQ3/6.  It appears that the overall code functionality will be adequate to meet the future anticipated needs of EQ3/6, which has been used to replace FMT for geochemical modeling.  
Design Phase 	

The DD, produced during the design phase, normally provides the following information: 

 Theoretical basis (physical process represented)
 Mathematical model (numerical model)
 Control flow and logic
 Data structures
 Functionalities and interfaces of objects, components, functions, and subroutines
 Ranges for data inputs and outputs in a manner that can be implemented into software
 
EQ3/6 is not a new computer code, nor was it specifically developed for application at WIPP; therefore, a DD is not required.  Earlier analyses for the WIPP project were done using FMT, which was also used to determine the actinide speciation and solubility for the CRA-2009 PA calculations.  Although DOE intends to replace FMT with EQ3/6, after careful benchmarking analyses and comparisons, PA-specific analyses to be conducted with EQ3/6 have not been specified.  A DD for EQ3/6 has not been developed; however, a number of the topics that are generally covered as part of the DD are included in the EQ3/6 UM (Wolery and Jarek 2003). 

Implementation Phase 

The UM and ID are produced during the implementation phase: 

User's Manual (UM)  -  describes the code's purpose and function, mathematical governing equations, model assumptions, the user's interaction with the code, and the models and methods employed by the code.  The UM generally includes the following:

 The numerical solution strategy and computational sequence, including program flowcharts and block diagrams.
 The relationship between the numerical strategy and the mathematical strategy (i.e., how boundary or initial conditions are introduced).
 A clear explanation of model derivation.  The derivation starts from generally accepted principles and scientifically proven theories.  The UM justifies each step in the derivation, and notes the introduction of assumptions and limitations.  For empirical and semi-empirical models, the documentation describes how experimental data are used to arrive at the final form of the models.  The UM clearly states the final mathematical form of the model and its application in the computer code.
 Descriptions of any numerical method used in the model that goes beyond simple algebra (e.g., finite-difference, Simpson's rule, cubic splines, Newton-Raphson Methods, and Jacobian Methods).  The UM explains the implementation of these methods in the computer code in sufficient detail so that an independent reviewer can understand them.
 The derivation of the numerical procedure from the mathematical component model.  The UM gives references for all numerical methods.  It explains the final form of the numerical model and its algorithms.  If the numerical model produces only an intermediate result, such as terms in a large set of linear equations that are later solved by another numerical model, then the UM explains how the model uses intermediate results.  The documentation also indicates those variables that are input to and output from the component model.

The Agency reviewed the UM for Version 8.0a (Wolery and Jarek 2003).  This document describes the functionality of EQ3/6, provides a code overview, describes material and element properties, and tells the analyst how to run EQ3/6.  The UM document requirements of NP 19-1 are satisfied by the UM (Table 6).  

Table  6.	NP 19-1 Criteria Matrix for User's Manual
                              NP 19-1 UM Criteria
    User's Manual for EQ3/6 Version 8.0a Section that Satisfies Criteria
1.	Functional Requirements and System Limitations
The functional requirements for EQ3/6 are described in Requirements Document for EQ3/6 Version 8.0a.
EQ3/6 Version 8.0a runs on a PC-compatible with Microsoft Windows 95, 98, 2000, NT, or XP; EQ3/6 may operate on other Windows systems.
2.	Mathematical model and numerical models
Section 2.1 (Conceptual Background),
Appendix B (EQ3 Supplemental Information),
Appendix D (EQ6 Supplemental Information), and
Appendix E (ODE Integration Methods)
3.	Physical and mathematical assumptions
Section 2.1 (Conceptual Background),
Appendix B (EQ3 Supplemental Information),
Appendix D (EQ6 Supplemental Information), and
Appendix E (ODE Integration Methods)
4.	Capabilities and limitations
Section 5 (Allowable/Tolerable Ranges for Inputs and Outputs)
5.	User's interaction
Run instructions are provided in Section 2.2 (Operating the Software), and in the VVP/VD for EQ3/6 Version 8.0a.
Section 2 (User Interactions) provides an overview of the interactions of the individual programs in the EQ3/6 package and user's interaction needed to run the codes.
6.	Input parameters, formats, valid ranges
Section 3.1 (EQ3/6 Input Database Files),
Section 3.2 (EQ3 and EQ6 Input Files),
Section 3.3 (The EQ3 Input File),
Section 3.4 (The EQ6 Input File),
Section 4 (File Formats), and
Section 5 (Allowable/Tolerable Ranges for Inputs and Outputs)
7.	Messages initiated by improper input
Section 3.2.2 (Input File Processing and Error Control) and
Section 6 (Anticipated Errors and User Response)
8.	Output specs and formats
Section 3.5 (EQ3 Output Files),
Section 3.6 (EQ6 Output Files), and
Section 3.7 (EQ6 Tabulated ("Tab") Output File)
9.	Required training
Section 8 (Required Training)
10.	Components not tested
The components not tested are defined in the VVP/VD for EQ3/6 Version 8.0a.


Implementation Document (ID)  -  provides the information necessary for the re-creation of the code used in the WIPP PA calculations.  Using this information, the computer user can reconstruct the code or install it on an identical platform to that used in the WIPP PA calculations.  The document includes the source-code listing, the subroutine-call hierarchy, and code compilation information.

Since EQ3/6 is not in the public domain, and DOE does not have a copy of the source code, DOE has not developed an ID.  Instructions for installing EQ3NR (Version 8) onto a PC are provided in Appendix A of the VVP/VD, however.  If the DOE does replace FMT with EQ3/6, it will be the only major PA code in which the source code is not readily available to the PA investigators.  

Validation Phase 

The validation phase consists of executing and reviewing test cases identified in the VVP to demonstrate that the developed software meets the requirements defined for it in the RD.  The VD, produced during this phase, summarizes the results of the testing activities prescribed in the RD and VVP documents for the individual codes and provides evaluations based on those results.  The VD contains listings of sample input and output files from computer runs of a model.  The VD also contains reports on code verification, bench marking, and validation, and documents results of the QA procedures. 

Requirements are outlined in Section 2.0 of the RD for EQ3/6 (Version 8.0a).  The 16 functional requirements described in Section 2.0 are those necessary for PA code usage.  These include the following requirements:

Functional Requirements

EQ3/6 is required to perform the following functions:

R.1	Perform aqueous speciation calculations, given total concentrations of dissolved components and other parameters, such as pH, pH, Cl, Eh, pe, oxygen fugacity, and CO2 fugacity.

R.2	Perform aqueous speciation calculations with charge balancing on a specified ion.

R.3 	Perform aqueous speciation calculations with mineral equilibrium constraints.

R.4 	Perform "single point" thermodynamic equilibrium calculations.

R.5	Perform reaction-path calculations without inclusion of chemical kinetics.

R.6	Perform reaction-path calculations with inclusion of chemical kinetics.

R.7	Perform reaction-path calculations for fluid-center flow-through open system.

R.8	Determine activity coefficients using Pitzer's equations, assuming an appropriate Pitzer thermodynamic data file is provided.

Performance Requirements

There are no performance requirements for EQ3/6.

Attribute Requirements

There are no attribute requirements for EQ3/6.

External Interface Requirements

R.9	EQ3NR and EQ6 require a binary thermodynamic data file.

R.10	EQ3NR requires a text input file (.3i) describing the speciation-solubility problem.

R.11	EQ3NR generates a text output file (.3o) describing the results of the calculation.

R.12	EQ3NR generates a text "pickup" file (.3p) that contains a compact description of the aqueous solution.  It may be used as the bottom part of an EQ6 input file.

R.13 	 EQ6 requires a text input file (.6i) describing the reaction-path problem.

R.14 	EQ6 generates a text output file (.6o) describing the results of the calculation.

R.15 	EQ6 generates a text "tab" file (.6t) that contains certain data in tabular form suitable for supporting local graphics post-processing.

R.16 	EQ6 generates a text "pickup" file (.6p) that may be used as an input file to restart a reaction path calculation where a previous run segment ended.

Other Requirements

The ability of the software to meet requirements R.1 through R.16 has already been established (SNL 2006a).  The present requirements document adds a new one:

R.17	EQ3/6, using an appropriate translation of the FMT database used in the WIPP geochemistry model, must produce results for WIPP-relevant and near-relevant problems that are substantially the same as those produced by FMT.  The WIPP-relevant problems must include examples involving actinides, and some must include both actinides and organic complexing agents.

All testing was performed on a PC with AMD Athlon(TM) Processor, AT/AT Compatible with Microsoft Windows XP Professional Version 2002, Service Pack 2.

The test set for EQ3/6 consists of 19 test cases, identified as Test Case #1 through Test Case #19.
The cases have been designed to verify that EQ3/6 Version 8.0a satisfies the requirements specified in Section 2 of the RD.  Test cases are summarized in Table 3.  Each test is discussed in detail in Section 6 of the VVP/VD.  The requirements covered by each test case were shown previously in Table 4.  Most test cases are designed to specifically verify a limited set of requirements, although many of the requirements are implemented by each test case.  

The description of test cases, input files, and acceptance criteria described in the VVP/VD exercise all code attributes required in the list provided in Section 2.0 of the RD.  In general, the techniques used to verify EQ3/6 are comparisons with independent codes with similar capabilities. 

The different techniques typically align with the nature of the test case -- individual component, or small groups of components.  This approach is a conventional way of validating computer codes, i.e., by a series of comparisons with known solutions that test various combinations of code options.  Table 7 provides a summary of where the VVP/VD NP-19-1 criteria are presented. 

Table 7.	NP 19-1 Criteria Matrix for the VVP/VD
                            NP 19-1 VVP/VD Criteria
                  EQ3/6 Report section that satisfies criteria
 Code Requirements
Functional; Performance; Attribute; External Interface; and Other Requirements pp. 7 - 8 (SNL 2006b)
 Code Functionality
pg 8 (SNL 2006b)
 Functional Testing
pp 9-59 (SNL 2006b)
 Installation and Regression Testing
pp 7-29 (SNL 2007)
 Input files
Appendix A (SNL 2007)
 Run Scripts
Appendix A (SNL 2007)

The Agency found that the EQ3/6 results satisfactorily met the acceptance criteria for all of the functional tests.  Although none of the EQ3/6 tests were designed expressly to test WIPP-specific geochemical conditions, many of the tests do encompass conditions that are similar to WIPP (e.g., high salinity).
                                       
Installation and Checkout Phase 

The following documents are produced during the installation and checkout (I&C) phase: 

 The Installation and Checkout (I&C) Form NP 19-1-8
 The Access Control Memorandum 
 The Approved Users Memorandum

The I&C Form for EQ3/6 Version 8 provides access control information and a list of approved users (SNL 2006e).  SNL maintains a series of software codes for application at the WIPP site.  Each code must be validated prior to initial use on each new platform.  The EQ3/6 code was acquired from LLNL.  The validation of EQ3/6 Version 8.0a was conducted in 2009 on a PC with Windows XP and documented in the VVP/VD. 

Regression testing must be conducted on each individual PC that runs EQ3/6 Version 8.0a for official WIPP work.  The I&C document for EQ3/6 covers the regression testing conducted to demonstrate the validity of EQ3/6 Version 8.0a on all "target" PCs listed below: 

S843806 - Intel(R) Pentium(R) 4 CPU 2.80GHz 
	Hardware Platform:  Intel Pentium 4 CPU 2.80GHz (S843806)
	Operating System:  Microsoft Windows XP Professional Version 2002 Service Pack 2

S838019 - Intel(R) Pentium(R) 4 CPU 2.20GHz 
	Hardware Platform:  Dell Workstation PWS340, Intel Pentium 4 CPU 2.20GHz (S8380I9)
	Operating System:  Microsoft Windows 2000 5.00.2195 Service Pack 4

S864333 - Intel(R) Xeon(TM) CPU 3.40GHz
	Hardware Platform:  Dell Workstation PWS670, Intel Xeonm CPU 3.40GHz (S864333)
	Operating System:  Microsoft Windows XP Professional Version 2002 Service Pack 2

S881469 - Intel(R) Pentium(R) D CPU 3.60GHz
	Hardware Platform:  Intel Pentium D CPU 3.60GHz (S88 1469)
	Operating System:  Microsoft Windows XP Professional Version 2002 Service Pack 2

Before testing, EQ3/6 must be installed on the target PC.  Installation instructions are provided in Appendix A of the VVP/VD.  The software package contains the executable programs in folder Eq3-6v8.O\Bin.  The executable programs and interface codes used to run the program are listed on each line, as explained in Appendix B of the VVP/VD. 

The I&C document presents the regression test results for the EQ3/6 Version 8.0a code on all target PCs listed above.  The tests for this code comprise the 19 test cases described in the VVP/VD.  Test results from EQ3/6 Version 8.0a run on each target PC were compared to results from the validation test of EQ3/6 Version 8.0a described in the VVP/VD.

NP 19-1 states that differences in run dates and times and execution statistics are acceptable.  These are the only differences expected in the EQ3/6 output files.  If all differences are found to be acceptable, it follows that the output of EQ3/6 Version 8.0a meets the acceptance criteria specified in the VVP/VD.  If the output of EQ3/6 Version 8.0a on each target PC meets the acceptance criteria specified in the VVP/VD, EQ3/6 Version 8.0a will be considered to be validated on each target PC in which it was tested.  The comparison found no differences in the numerical output of EQ3/6 Version 8.0a on any target PC and the VVP/VD validation. 

The Agency concludes that EQ3/6 Version 8.0a meets the acceptance criteria in the VVP/VD and has passed the regression testing on all target PCs listed above.

Production Software and/or Baseline Document Change Control 

When changes to the software baseline occur, the Change Control Form, Form NP 19-1-9, is used.  The following are types of changes that may be implemented:

 Major changes, including new requirements, new design, new models, new implementation, require a new baseline (i.e., SQAP, RD, DD, VVP, ID, UM, VD) to be documented.  In addition to revising every baseline document, a Change Control Form and an I&C Form are used.

 Minor changes do not affect the requirements or design, and can be documented with addenda (no more than three addenda per baseline document) or page changes to the affected baseline document, in addition to the Change Control Form and the I&C Form. 

 Patch changes can be used for very small fixes to the code, usually one or two lines of source code or expanding a field's character length, etc.  Patch changes can be documented and tested with the Change Control Form and I&C Form. 
The Change Control Form for EQ3/6 Version 8.0 indicates that a number of changes have been made to the code since the release of Version 7.2c (SNL 2006f).  Version 8.0 of EQ3/6 represents a complete rewrite of the Version 7 series software, which dates back to the start of EQ3/6 at Northwestern University in the mid-1980s.  Because it is a complete rewrite, SNL notes that Version 8.0 may not be as stable as Version 7.2c, which benefited from years of feedback by code users.  On the other hand, Version 8.0 does contain several improvements over previous versions.  The latest software is mostly dynamically dimensioned, including everything with regard to the supporting data file.  A few items in the input files are still statically dimensioned, however, like the number of ion exchange phases.  The menu-style input file format has been simplified and should be easier to understand and use.  The ordinary differential equation (ODE) integrator has been expanded into a full predictor corrector method with stiff-system capability.  The stiff-system solver uses a true Jacobian matrix.  This should eliminate those cases where the step sizes become stuck at very small values, because the system of rate equations becomes intractable. 

Associated with this development are several enhancements including:

       The availability of a new option for treating surface areas (i.e., constant particle number/geometric growth). 
       A new high-temperature Pitzer modeling capability.
       An upgrade to the code numerics in dealing with extremely concentrated.
       The addition of a generic multi-exchanger, multi-site ion exchange model.
       Input file entries to specify exchanger phases and sites and associated thermodynamic data.
       The addition of high-pressure capability.  This is associated with the use of some data files (e.g., data0.shv, data0.500, data0.2kb, and data0.5kb) derived from SUPCRT92 and its own associated data files.  Each file has a reference pressure (the shv file follows the classic 1 atm steam saturation pressure curve), but EQ3/6 will extrapolate to other pressures as specified on the input file. 
       Specified redox disequilibria for reaction path runs. 
       Options for separating fluids from associated minerals and for facilitating fluid mixing calculations.

The Change Control Form also indicates that the VVP/VD contained an error in the tolerance of the log fugacity of CO2 in Test Case #2.  This tolerance was subsequently expanded to +-0.2.  No other changes to the VVP/VD were made and no other QA documents were changed.  No changes to the code were required.

The Change Control Form for EQ3/6 Version 8.0a indicates that a number of changes have been made to the code since the release of Version 8.0 (SNL 2009).  These revisions are presented in Section A.3.iv.

System Software and Hardware Change Control 

Coding Documentation Standards.  Any change to software must be accompanied by documentation describing the change, the date the change was made, and the name of the person responsible for implementing the change.  This documentation should be clearly identified and placed in the code in the vicinity of the change, as well as at the top of the code prior to the first executable line.  The code reviewer shall determine if this documentation is clear and sufficient. 

Significant System Software or Hardware Changes.  The Code Team/Sponsor (single-user systems) or System Administrator (multi-user systems) proposes significant system software or hardware changes using the Change Control Form NP 19-1-9.  Examples of significant changes to system software or hardware are as follows: 

 Changes to the operating system, such that the version or level identifier changes
 Changes to the Central Processing Unit (CPU) 
 Database management system change 
In general, changes are significant if they impact the results generated by production software or cause recompilation of production software. 

Since EPA's certification of the WIPP CCA, SNL has added computer hardware and upgraded the computer software.  In order to maintain compliance with 40 CFR 194.22 and 194.23, SNL is required to conduct regression testing on the computer codes to ensure that they still function properly on new hardware and software.

Tests #1 through #15 in the VVP/VD for Version 8.0a were tested with previous versions of EQ3/6.  DOE notes, however, that changes to the input and output files for these tests make standard "regression testing" difficult, so SNL has validated these tests on a case-by-case basis by comparing model predicted output values to experimental results or to values calculated by independent codes. 

Software Problem Report (SPR).  Whenever a software problem is identified, the Code Team/Sponsor evaluates the problem to determine if it is, indeed, a problem (as opposed to user error).  If it is a problem, the SPR process is followed. 

The Code Team/Sponsor classifies the problem as major if it could significantly impact previous uses of code, or if it will require significant modification to the software; otherwise it is classified as minor.  For major problems, the responsible manager identifies affected users to be notified of the problem, and designates qualified personnel to identify and evaluate the impact of the software problem.  The affected analyses are revised as necessary, and the evaluation and resolution of the software problem are documented in Part II of the SPR and Evaluation Form.  For minor problems, this evaluation can be performed by the Code Team/Sponsor.

No SPRs have been generated for EQ3/6 Version 8.0a.

Configuration Management (Configuration Identification and Status Accounting).  Configuration Management refers to the process for defining the configuration of software products, establishing software configuration baselines, and tracking the status of baseline changes.  A software configuration baseline consists of the source code and baseline documents, providing objective evidence of technical adequacy.

The Software Configuration Management (SCM) Coordinator maintains a Software Baseline List and makes it available upon request.  The SCM Coordinator performs a completeness review to ensure compliance with the procedure, and to ensure that necessary components of configuration management are present.  The Software Baseline List contains the following:

 Code name and version
 Code version date
 Code Team/Sponsor name
 Code classification
 RD version
 VVP version
 DD version
 ID version
 UM version
 VD version
 List of approved users (may be listed by name, organization, group, or task, etc.) 
 List of approved system software/hardware configurations
 List of outstanding SPR numbers
 Status of approved changes in process
 I&C (Installation and Checkout) date

Such documentation shall include, but shall not be limited to the following:

c.1) 	Description of the theoretical backgrounds of each model and the method of analysis or assessment,
Theoretical aspects of the model construction are described in Sections 2 and 3 of Wolery and Daveler (1992c) and Section 2 of Wolery (1992a) and are summarized below.

EQ3NR is an aqueous solution speciation-solubility modeling code.  It is part of the EQ3/6 software package for geochemical modeling.  It computes the thermodynamic state of an aqueous solution by determining the distribution of chemical species, including simple ions, ion pairs, and complexes, using standard state thermodynamic data and various equations that describe the thermodynamic activity coefficients of these species.  The input to the code describes the aqueous solution in terms of analytical data, including total (analytical) concentrations of dissolved components and such other parameters as the pH, Eh, pe, and oxygen fugacity.  The input may also include a desired electrical balancing adjustment and various constraints, which impose equilibrium with specified pure minerals, solid solution end-member components (of specified mole fractions), and gases (of specified fugacities).  The code evaluates the degree of disequilibrium in terms of the saturation index for various reactions, such as mineral dissolution or oxidation-reduction in the aqueous solution itself.  Individual values of Eh, pe, oxygen fugacity, and Ah (redox affinity) are computed for aqueous redox couples.  Equilibrium fugacities are computed for gas species.  The code is highly flexible in dealing with various parameters as either model inputs or outputs.  The user can specify modification or substitution of equilibrium constants at run time by using options on the input file.  The output consists of an output file and a pickup file, which can be used to initialize an EQ6 reaction path calculation.  EQ3NR may be used by itself or to initialize a reaction path calculation by EQ6, its companion code in the EQ3/6 package. 

EQ6 is a FORTRAN computer program in the EQ3/6 software package (Wolery 1979).  It calculates reaction paths (chemical evolution) in reacting water-rock and water-rock-waste systems.  Speciation in aqueous solution is an integral part of these calculations; EQ6 computes models of titration processes (including fluid mixing), irreversible reaction in closed systems, irreversible reaction in some simple kinds of open systems, and heating or cooling processes, as well as solving "single-point" thermodynamic equilibrium problems.  A reaction path calculation normally involves a sequence of thermodynamic equilibrium calculations.  Chemical evolution is driven by a set of irreversible reactions (i.e., reactions out of equilibrium) and/or changes in temperature and/or pressure.  These irreversible reactions usually represent the dissolution or precipitation of minerals or other solids.  The code computes the appearance and disappearance of phases in solubility equilibrium with the water.  It finds the identities of these phases automatically.  The user may specify which potential phases are allowed to form and which are not.  There is an option to fix the fugacities of specified gas species, simulating contact with a large external reservoir.  Rate laws for irreversible reactions may be either relative rates or actual rates.  EQ6 is able to model both mineral dissolution and growth kinetics.  The user can specify modification or substitution of equilibrium constants by using options on the input file.  The output consists of an output file, a tab file (tables of output parameters), and a pickup file, which allows a restart capability.  The chief numerical method employed for equilibrium calculations is a hybrid Newton-Raphson technique.  This is supported by a set of algorithms which create and optimize starting values.  When actual rate laws are used, EQ6 integrates the resulting equations using a finite difference-based ODE solver.  EQ6 reads a secondary unformatted data file by EQPT, the EQ3/6 data file preprocessor.  There is currently a set of five data files.  Three of these may be used with either the Davies equation or the B-dot equation to describe the activity coefficients, and their use is restricted to modeling dilute solutions.  The other two of these use Pitzer's equations and are suitable for modeling solutions to high concentrations, though with fewer chemical components.  The temperature range of the thermodynamic data in the data files varies from 0 to 100°C.  The companion code EQ3NR must be used to initialize a reaction path calculation by EQ6.  EQ6 and the other codes in the EQ3/6 package are written in FORTRAN 77 and were originally developed to run under the UNIX operating system on computers ranging from workstations to supercomputers.

            c.2)  	General descriptions of the models, discussions of the limits of the applicability of each model, detailed instructions for executing the computer codes, including hardware and software requirements, input and output formats and explanations of each input and output variable and parameter (e.g., parameter name and units); listings of input and output files from a sample computer run; and reports on code verification, bench marking, validation, and quality assurance procedures.

A general description of EQ3/6 Version 8.0a is provided in the UM, including detailed instructions for executing the code (Wolery and Jarek 2003).  Lists of variable names are also provided in Wolery and Daveler (1992c) and Wolery (1992b).  Hardware and software requirements, and input and output formats with explanations of variables are provided in the UM with additional information in the VVP/VD (Wolery and Daveler 1992c and Wolery 1992b).  A series of listings of example input and output files are given in the UM, and additional files are provided in Wolery and Daveler (1992c) and Wolery (1992b).  Reports on code verification and validation are provided in the VVP/VD.  The verification and validation study involved 18 test cases that cover the functional requirements of the EQ3/6 code.  The UM and VVP/VD have been subjected to internal peer review, as indicated by QA sign-off sheets attached to the references. 
 
Since the application of EQ3/6 to future CRAs has not yet been defined, potential limits of applicability for EQ3/6 will have to be evaluated on a case-by-case basis.  However, it is anticipated that EQ3/6 will be employed to simulate actinide chemistry.  

c.3)  	Detailed descriptions of the structure of computer codes and complete listings of the source codes,

The generalized structure of the file system and information flow for the EQ3/6 code is summarized in Figure 1 of the UM (Wolery and Jarek 2003).  EQ3/6 is written in FORTRAN 77.  The hardware platform for EQ3/6 is any personal computer running a 32-bit Windows operating system (e.g., Windows 95 and onward).  Support for UNIX computers has been discontinued.  Hard disk drive space requirements will depend on the users' typical output file size.  EQ3NR output files (extension .3o) are quite small, typically <100 KB, whereas EQ6 output files (extension .6o) range from a typical 1 MB to tens of MBs, depending upon the time (reaction path) and printout intervals.  

The EQ3/6 code was acquired from LLNL, and SNL does not have the source code. 

c.4) 	Detailed descriptions of data collection procedures, sources of data, data reduction and analysis, and code input parameter development,

The EQ3/6 code depends on a database of thermodynamic constants characteristic of specific reactions.  Chemical potentials and Pitzer ion interaction parameters for the major elements are included in the EQ3/6 database.  Presumably, the EQ3/6 database will be similar to that used by FMT.  Constants for the actinides relevant to specific reactions described in Appendix SOTERM [CRA-2004] are taken from a variety of sources published in the scientific peer-reviewed literature.  As described in Appendix SOTERM, constants for the actinides were generated from fits to solubility data using the NONLIN program (Docket A-93-02, II-G-1, Ref. #477, and SNL 1996, cited in Appendix SOTERM).  The NONLIN program contains the same database of Pitzer ion interaction parameters and chemical potentials as used by EQ3/6, so that internal consistency is maintained in the generation of new constants for inclusion in the EQ3/6 database.  Details on the use of the NONLIN code are not provided in Appendix SOTERM of the CCA, but the reader is referred to the references above.  Although Appendix SOTERM identifies specific sources of thermodynamic data for the FMT code, it does not provide an indication of the specific database used for the PA calculations presented in the CCA.  Supplementary references cited in Appendix SOTERM also do not indicate which thermodynamic database was used for model calculations of actinide concentrations for the PA as presented in Table SOTERM-2 of Appendix SOTERM.  The database has undergone continuous revision since the CCA, which has probably resulted in improvements in its values and completeness, but has apparently obscured the identity of the database version specifically used for the CCA.  This issue is raised because the results of chemical equilibrium problems are primarily dependent on the thermodynamic data used to represent pertinent reactions.  As noted earlier, a complete copy of the database is included in Novak 1997, and this updated version was included in the EPA mandated PAVT (Docket A-93-02, II-G-26).

      c.5) 	Any necessary licenses,

The EQ3/6 code was acquired and licensed from LLNL.

            c.6)  	An explanation of the manner in which models and computer codes incorporate the effects of parameter correlation.

The effects of parameter correlation are not explicitly discussed in the UM and VVP/VD.  However, it is generally understood that equilibrium reactions result in nonlinear correlations between concentrations and master variables, such as pH or CO2(g) fugacity, because of various hydrolysis and complexation reactions.  These effects are explicitly accounted for in the reaction
stoichiometries in the  EQ3/6 thermodynamic database and related constants.
      d) 	The Administrator or the Administrator's authorized representative may verify results of computer simulations used to support any compliance application by performing independent simulations.  Data files, source codes, executable versions of computer software for each model, other material or information needed to permit the Administrator or the Administrator's authorized representative to perform independent simulations, and access to necessary hardware to perform such simulations, shall be provided within 30 calendar days of a request by the Administrator or the Administrator's authorized representative.

The Agency may elect to conduct independent evaluations of EQ3/6 at a later date.


EPA Findings

EPA reviewed all of the relevant documentation pertaining to EQ3/6 Version 8.0a as described above (e.g., UM, VVP/VD, and VD), as well as supplementary information provided by SNL.  The test cases summarized by SNL (2011) adequately represent calculations performed for WIPP PA and results of the test case calculations show that the EQ3/6 performance is acceptable.  However, as summarized in Section a.3.iv, some minor errors identified in the report should be corrected.  

With the exception of the source code not being readily available, EPA found that SNL's QA requirements for PA and compliance assessment are in agreement with those specified in § 194.22, and that the code documentation for EQ3/6 is adequate and meets the requirements of  40 CFR Part 194.  SNL's documentation includes plan(s) for QA software, software requirements documentation, software design and implementation documentation, software verification and validation documentation, and user documentation.

References

ASME (American Society of Mechanical Engineers) 1990.  NQA-2a-1990 addenda, part 2.7, to ASME NQA-2-1989 edition, "Quality Assurance Requirements for Nuclear Facility Applications;" IBR approved for Sec. 194.22 and Sec. 194.23.

Babb, S.C. and C.F. Novak, 1997.  User's Manual for FMT Version 2.3:  A Computer Code Employing the Pitzer Activity Coefficient Formalism for Calculating Thermodynamic Equilibrium in Geochemical Systems to High Electrolyte Concentrations.  Document Version 2.3, Sandia National Laboratories, WPO # 43037. 
Brush, L.H., Y. Xiong, J.W. Garner, A. Ismail, and G.T. Roselle, 2006.  Consumption of Carbon Dioxide by Precipitation of Carbonate Minerals Resulting from Dissolution of Sulfate Minerals in the Salado Formation in Response to Microbial Sulfate Reduction in the WIPP.  Sandia National Laboratories, Carlsbad, New Mexico.
Brush, L.H., Y. Xiong and J.J. Long, 2009.  Results of the Calculations of Actinide Solubilities for the WIPP CRA-2009 PABC.  Sandia National Laboratories, Carlsbad, New Mexico, October 2009, ERMS 552201.

Bynum, R.V., 1996.  "Analysis to Estimate the Uncertainty for Predicted Actinide Solubilities."  WBS 1.1.10.1.1, Rev. 0, effective date 9/3/96 (WPO41374).
Chavez, M.J., 2011.  Re:  Input and Output File Request.  E-mail to J.A. Schramke (SC&A) from M.J. Chavez (SNL), February 9, 2011.
Clegg, S.L. and P. Brimblecombe, 1990.  "The solubility and activity coefficient of oxygen in salt solutions."  Geochimica Cosmochimica Acta 54:3315-3328.
DOE (U.S. Department of Energy) 1996.  Title 40 CFR Part 191 Compliance Certification Application for the Waste Isolation Pilot Plant, DOE/CAO-1996-2184, October 1996, Carlsbad Field Office, Carlsbad, New Mexico.

DOE (U.S. Department of Energy) 2004.  Title 40 CFR 191 Parts B and C Compliance Recertification Application, U.S. Department of Energy Field Office, March 2004.  Docket A-98-49 Category II-B2.
EPA (U.S. Environmental Protection Agency) 2006.  Technical Support Document for Section 194.24:  Evaluation of the Compliance Recertification Actinide Source Term and Culebra Dolomite Distribution Coefficient Values.  Office of Radiation and Indoor Air, Docket No: A-98-49, March 2006.

Fanghänel, T., and J.I. Kim, 1998.  "Spectroscopic evaluation of thermodynamics of trivalent actinides in brines."  Journal of Alloys and Compounds 271-273:728-737. 

Fanghänel, T., V. Neck, and J.I. Kim, 1995.  "Thermodynamics of neptunium(V) in concentrated salt solutions: II. Ion interaction (Pitzer) Parameters for Np(V) hydrolysis species and carbonate complexes."  Radiochimica Acta 69:169-176.

Felmy, A.R., and J.H. Weare, 1986.  "The prediction of borate mineral equilibria in natural waters: Application to Searles Lake, California."  Geochimica et Cosmochimica Acta 50:2771-2783.

Felmy, A.R., D. Rai and M.J. Mason, 1991.  "The solubility of hydrous thorium(IV) oxide in chloride media: development of an aqueous ion-interaction model."  Radiochimica Acta 55:177-185. 

Harvie, C.E., 1981.  "Theoretical Investigations in Geochemistry and Atom Surface Scattering."  PhD Dissertation, University of California, San Diego.
Harvie, C.E., N. Møller, and J.H. Weare, 1984.  "The prediction of mineral solubilities in natural waters:  The Na-K-Mg-Ca-H-Cl-SO4-OH-HCO3-CO3-CO2-H2O system to high ionic strengths at 25°C."  Geochimica et Cosmochimica Acta 48:723-751.
IEEE (Institute of Electrical and Electronics Engineers) 1986.  IEEE Standard for Software Verification and Validation Plans, pp. 1012 - 1986, New York, New York:  IEEE Press.

Ismail, A.E., H. Deng, J.-H. Jang and T.J. Wolery, 2009.  Verification of FMT Database and Conversion to EQ3/6 Format.  Sandia National Laboratories, Carlsbad, New Mexico, January 13, 2009, ERMS 550689.

Jenne, E.A., 1981.  Geochemical Modeling:  A Review.  PNL-3574, Battelle Pacific Northwest Laboratory), Richland, Washington.

Lide, D.R., Ed., 1999.  CRC Handbook of Chemistry and Physics.  CRC Press, New York.

Novak, C.F., 1997.  "Calculation of Actinide Solubilities in WIPP SPC and ERDA6 Brines Under MgO Backfill Scenarios Containing Either Nesquehonite or Hydromagnesite as the Mg-CO3- Solubility-Limiting Phase.  Memorandum to R.V. Bynum, April 21, 1997.  WPO 38628, Sandia National Laboratories, Albuquerque, New Mexico.

Novak, C.F., I. Al Mahamid, K.A. Becraft, S.A. Carpenter, N. Hakem, and T. Prussin, 1996.  "Measurement and Thermodynamic Modeling of Np(V) Solubility in Dilute through Concentrated K2CO3 Media."  SAND96-1604J, Sandia National Laboratory, Albuquerque, New Mexico.

Novak, C.F., N.J. Dhooge, H.W. Papenguth, and R.F. Weiner, 1995.  "Systems Prioritization Method  -  Iteration 2 Baseline Position Paper:  Actinide Source Term."  Sandia National Laboratory, Albuquerque, New Mexico, dated March 31, 1995.

Pitzer, K.S., 1975.  Thermodynamics of electrolytes.  "V. Effects of higher-order electrostatic terms."  Journal of Solution Chemistry 4:249-265.
Pitzer, K.S., 1991.  "Ion interaction approach: theory and data correlation," Chapter 3 in Activity Coefficients in Electrolyte Solutions.  Ed. K.S. Pitzer.  CRC Press, 2[nd] Edition.

Plummer, L.N., D.L. Parkhurst, G.W. Fleming and S.A. Dunkle, 1988.  A Computer Program Incorporating Pitzer's Equations for Calculation of Geochemical Reactions in Brines.  U.S. Geological Survey, Water-Resources Investigations Report 88-4153, Reston, Virginia.
Rai, D., A.R. Felmy, S.M. Sterner, D.A. Moore and M.J. Mason, 1997.  "The solubility of Th(IV) and U(IV) hydrous oxides in concentrated NaCl and MgCl2 solutions."  Radiochimica Acta 79:239-247.
SC&A (S. Cohen and Associates) 2008.  "Review of MgO-Related Uncertainties in the Waste Isolation Pilot Plant."  SC&A, Inc., Vienna, Virginia.

Shock, E.L. and H.C. Helgeson, 1988.  "Calculation of the thermodynamic and transport properties of aqueous species at high pressures and temperatures: correlation algorithms for ionic species and equation of state predictions to 5 kb and 1000°C."  Geochimica Cosmochimica Acta 52: 2009-2016. 

Shock, E.L., H.C. Helgeson, and D.A. Sverjensky, 1989.  "Calculation of the thermodynamic and transport properties of aqueous species at high pressures and temperatures: Standard partial molal properties of inorganic neutral species."  Geochimica et Cosmochimica Acta 53:2157-2183.

Shoemaker, P.E., 2011. Requested EQ3NR files. Letter to J.A. Schramke, July 21, 2011. 

SNL (Sandia National Laboratories) 1996.  "Quality Assurance Document for EQ3/6 (Version 7.2a,b)."  Sandia National Laboratories.  Sandia WIPP Central Files ERMS #24125.

SNL (Sandia National Laboratories) 2004.  "Nuclear Waste Management Procedure NP 19-1 Software Requirements," Rev. 11, Sandia National Laboratories, Albuquerque, New Mexico, August 2004.

SNL (Sandia National Laboratories) 2006a.  "Verification and Validation Plan/Validation Document for EQ3/6 Version 8.0."  Sandia National Laboratories, Albuquerque, New Mexico.  (ERMS# 543603)

SNL (Sandia National Laboratories) 2006b.  "Verification and Validation Plan/Validation Document for EQ3/6 Version 8.01."  Sandia National Laboratories, Albuquerque, New Mexico.  (ERMS# 544574)

SNL (Sandia National Laboratories) 2006c.  "Requirements Document for EQ3/6 (Version 8.0)."  Sandia National Laboratories.  Sandia WIPP Central Files ERMS #543602.

SNL (Sandia National Laboratories) 2006d.  "Quality Assurance Plan for EQ3/6 (Version 8.0)."  Sandia National Laboratories.  Sandia WIPP Central Files ERMS #543601.

SNL (Sandia National Laboratories) 2006e.  "Software Installation and Checkout for EQ3/6, Version 8.0."  Sandia National Laboratories, Albuquerque, New Mexico.  (ERMS# 544310)
SNL (Sandia National Laboratories) 2006f.  "Change Control for EQ3/6, Version 8.0."  Sandia National Laboratories, Albuquerque, New Mexico.  (ERMS# 544614)

SNL (Sandia National Laboratories) 2007.  "Software Installation and Checkout for EQ3/6, Version 8.0."  Regression Testing for Windows PCs.  Sandia National Laboratories, Albuquerque, New Mexico.  (ERMS# 545285)

SNL (Sandia National Laboratories) 2009.  "Change Control for EQ3/6, Version 8.0a."  Sandia National Laboratories, Albuquerque, New Mexico.  (ERMS# 551823)

SNL (Sandia National Laboratories) 2011.  "Verification and Validation Plan/Validation Document for EQ3/6 Version 8.0a for Actinide Chemistry, Revision 1."  Document Version 8.2020, Carlsbad, New Mexico, May 12, 2011. (ERMS #555358)

TEA (Trinity Engineering Associates) 2004.  Review of Effects of Supercompacted Waste and Heterogeneous Waste Emplacement on WIPP Repository Performance.  Final Report.  Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC, March 17, 2004.

Wilson, C., D. Porter, J. Gibbons, E. Oswald, G. Sjoblom, and F. Caporuscio, 1996.  Conceptual Models Peer Review Report, Prepared for the U.S. Department of Energy, Carlsbad, New Mexico, July 1996, Docket No. A-93-02 Item II-G-1.
Wolery, T. J., 1979.  "Calculation of Chemical Equilibrium between Aqueous Solutions and Minerals:  The EQ3/6 Software Package."  UCRL-52658.  Lawrence Livermore National Laboratory, Livermore, California.

Wolery, T.J., 1992a.  "EQ3/6, A Software Package for Geochemical Modeling of Aqueous Systems:  Package Overview and Installation Guide (Version 7.0)."  UCRL-MA- 1 10662-PT-I.  Lawrence Livermore National Laboratory, Livermore, California. 

Wolery, T.J., 1992b.  "EQ3, A Computer Program for Geochemical Aqueous Speciation- Solubility Calculations:  Theoretical Manual, User's Guide, and Related Documentation (Version 7.0)."  UCRL-MA-110662-PT-III.  Lawrence Livermore National Laboratory, Livermore, California.

Wolery, T.J., 2008.  "Analysis Plan for EQ3/6 Analytical Studies, Task 1.4.1.1."  AP-140, Rev. 0.  Sandia National Laboratories, Carlsbad, New Mexico.

Wolery, T.J. and S.A. Daveler, 1992a.  "EQPT, A Data File Preprocessor for the EQ3/6
Software Package:  User's Guide and Related Documentation (Version 7.0)."  UCRL-MA-110662-PT-I.  Lawrence Livermore National Laboratory, Livermore, California

Wolery, T.J. and S.A. Daveler, 1992b.  "EQPT, A Data File Preprocessor for the EQ3/6 Software Package:  User's Guide and Related Documentation (Version 7.0)."  UCRL-MA-110662-PT-II.  Lawrence Livermore National Laboratory, Livermore, California.

Wolery, T.J. and S.A. Daveler, 1992c.  "EQ6, A Computer Program for Reaction Path Modeling of Aqueous Geochemical Systems:  Theoretical Manual, User's Guide, and Related Documentation (Version 7.0)."  UCRL-MA-110662-PT-IV.  Lawrence Livermore National Laboratory, Livermore, California.
Wolery, T.J. and R.L. Jarek, 2003.  "Software User's Manual for EQ3/6, Version 8.0."  U.S. Department of Energy, Office of Civilian Radioactive Waste Management, Las Vegas, Nevada.  Software Document Number 108 13-UM-8.0-00.

Xiong, Y., 2005.  Release of FMT_050405.CHEMDAT.  E-mail to J.F. Kanney and J.J. Long, April 5, 2005.  Sandia National Laboratories, Carlsbad, New Mexico, ERMS 539304.

Xiong, Y., L.H. Brush, A.E. Ismail and J.J. Long, 2009.  Uncertainty Analysis of Actinide Solubilities for the WIPP CRA-2009 PABC.  Sandia National Laboratories, Carlsbad, New Mexico, ERMS 552500.


 ATTACHMENT A:  DATABASE MIGRATION

The FMT database used in the PABC09 actinide solubility calculations was FMT_050405.CHEMDAT (Brush et al. 2009, Xiong 2005).  Ismail et al. (2009) describes the conversion of the FMT database FMT_050405.CHEMDAT to a format compatible with EQ3/6 (data0.fmt.R0).  The data included in the EQ3/6 database includes atomic weights for the elements in the database; log K data for aqueous species, gases and solid phases; and Pitzer parameters.

  Atomic Weights

Atomic weights are included in the EQ3/6 database to allow the conversion of input concentration units, for example mg/L, to units of molality (moles/kg solvent H2O) used in EQ3/6 calculations.  The source of the atomic weights cited in data0.fmt.R0 was "99lid via data0.ymp.R5." Neither data0.ymp.R5 nor data0.fmt.R0 provides information regarding this reference, although it is likely Lide (1999).  Information regarding the "99lid" reference should be included in the EQ3/6 database.

Atomic weights are included in data0.fmt.R0 for the elements listed in Table A-1.  FMT_050405.CHEMDAT provides atomic weight data for the elements in Table A-1 and for PosIon, NegIon, oxalate [C2O4[2-]], acetate [C2H3O2[-]], perchlorate [ClO4], lactate [C3H5O3[-]], EDTA [C10H12O8N2[4][-]] and citrate [C6H5O7[3-]].  PosIon and NegIon are pseudo-species included in the FMT database to permit charge balancing of the modeled solutions.  The pseudo-elements Null- and Null+ were included in data0.fmt.R0 with atomic weights of 1  10[-38] g/mole and the aqueous species NegIon and PosIon were assumed to be composed of the fictitious elements Null- and Null+, respectively.  Oxalate, acetate, perchlorate, lactate, EDTA and citrate were included in data0.fmt.R0 as auxiliary basis species.  Because molecular weights of the auxiliary basis species are calculated by EQ3/6 using chemical formulas and the atomic weights provided for the elements in Table A-1, molecular weights for oxalate, acetate, perchlorate, lactate, EDTA and citrate are not included in data0.fmt.R0.

                 Table A-1.	Elements Included in Data0.fmt.R0
Oxygen
Chloride
Neptunium
Americium
Hydrogen
Phosphorus
Boron
Potassium
Sulfur
Bromine
Magnesium
Thorium
Carbon
Nitrogen
Null-
Calcium
Sodium
Null+

The atomic weights listed for hydrogen, chlorine, boron and phosphorus in data0.fmt.R0 differ slightly from the values in FMT_050405.CHEMDAT because of an additional significant figure in the data0.fmt.R0 database.  The atomic weight of 32.06 g/mole for sulfur listed in FMT_050405.CHEMDAT differs slightly from the value of 32.066 g/mole included in data0.fmt.R0.  The atomic weights listed in data0.fmt.R0 are correct, based on a comparison to Lide (1999).

The atomic weight data in FMT_050405.CHEMDAT have had no effect on past WIPP PA calculations because all data were input in units of molarity (moles/L).  Similarly, the atomic weight data in data0.fmt.R0 are unlikely to affect future WIPP PA calculations performed using EQ3/6, because input data are likely to be in units of molarity or molality (moles/kg solvent H2O), which will not require use of the atomic weight data to obtain the molality units used in the EQ3/6 calculations.

 Strict Basis Species

EQ3/6 divides aqueous species in the database into three categories: strict basis species, auxiliary basis species and other aqueous species.  Each strict basis species represents a chemical element, with the exception of the O2(g) species that serves as the redox species.  The strict basis species included in data0.fmt.R0 are listed in Table A-2.  The information included in the EQ3/6 database for each strict basis species are the charge and the elemental composition of the species.  These data, which are checked by EQ3/6 during preprocessing, were visually examined and found to be complete and accurate.

                Table A-2.	Strict Basis Species in Data0.fmt.R0
H2O
Cl[-]
NpO2[+]
Am[3+]
H[+]
HPO4[2-]
B(OH)4[-]
K[+]
SO4[2-]
Br-
Mg[2+]
Th[4+]
HCO3[-]
NO3[-]
NegIon
Ca[2+]
Na[+]
PosIon

 Auxiliary Basis Species

Each auxiliary basis species represents an aqueous species that can be modeled to persist in solution without reacting to form its constituent elements.  Data included in data0.fmt.R0 for the auxiliary basis species include the charge, elemental composition, the identity and stoichiometry of the dissociation reaction, the associated log K for the reaction and the source of the log K data.

Oxalate is an example of an auxiliary basis species included in WIPP PA calculations.  Oxalate can react to form bicarbonate [HCO3[-]]:

                     C2O4[2-] + H2O + (1/2) O2(g)  2 HCO3[-]

Because WIPP PA calculations include the assumption that the organic ligands, including oxalate, acetate, citrate, EDTA and lactate, are neither consumed nor produced during the 10,000 year regulatory time period, the organic ligands are included as auxiliary basis species in data0.fmt.R0.  Because organic ligands are not modeled to be either consumed or produced, no log K data are required or provided for the organic ligands in the data0.fmt.R0 database.

The other species included as auxiliary basis species in data0.fmt.R0 are O2(aq), H2(aq), NH3(aq) and perchlorate.  The species O2(aq), H2(aq) and NH3(aq) are not included in the FMT_050405.CHEMDAT database, but were added to data0.fmt.R0.  Ismail et al. (2009) stated that the log K data for O2(aq) and H2(aq), originally derived from data reported by Shock et al. (1989), were obtained from data0.ymp.R5.  However, the cited reference in the data0.fmt.R0 database for these species is data0.ypf.R1.  The log K data for O2(aq) and H2(aq) are the same in all three databases (data0.fmt.R0, data0.ymp.R5 and data0.ypf.R1).  Citations for the original data source (Shock et al. 1989) should be included in data0.fmt.R0.

The source of the 25°C log K data for NH3(aq) is cited in data0.fmt.R0 as "DBConversion050405.xls."  Examination of the spreadsheet DBConversion050405_Rev3.xls cited by Ismail et al. (2009) showed that the standard chemical potentials (u[0]/RT) of NH3(aq) and NO3[-] were calculated from the free energy of formation data (ΔGf[0]) in data0.ymp.R5.  These data and standard chemical potential data for O2(g) (zero by convention) and for H[+] and H2O from Harvie et al. (1984) were used to calculate the log K value for the reaction as described below (Section A.4):

                    NH3(aq) + 2 O2(g) = NO3[-] + H[+] + H2O

The accuracy of the log K value for this reaction in data0.fmt.R0 was confirmed by independently performing the calculation.  The source citation should be corrected from ""DBConversion050405.xls" to "DBConversion050405_Rev3.xls" and the original sources of the free energy of formation and standard chemical potential data used in the calculation (Harvie et al. 1984, Shock and Helgeson 1988, Shock et al. 1989) should also be included in the source citation.

The standard chemical potential for perchlorate was included in the FMT database, but the FMT database did not include data for the O2(g) species required to calculate the log K for the reaction:
                                       
                          ClO4[-] = Cl[-] + 2 O2(g) 

Consequently, the log K data for perchlorate were obtained from data0.ymp.R5.  Examination of data0.ymp.R5 showed that the data were obtained from this source.  However, the original data source for the log K values was not provided; an explanation of the original source of the perchlorate log K data in data0.ymp.R5 should be provided.

 Other Aqueous Species and Solid Phases

Data provided in the EQ3/6 database for aqueous species that are not strict basis species or auxiliary basis species include the charge, elemental composition, the identity and stoichiometry of the dissociation reaction, log K values for the reaction and the source of the log K values.  For solid phases, the data in the EQ3/6 database include the elemental composition, the identity and stoichiometry of the dissolution reaction, log K values for the reaction and the source of the log K values.  A field is also provided for the molar volume (V0PrTr), but because this parameter is unused in WIPP calculations, it was set equal to zero for all solid phases (Ismail et al. 2009).

The data for the aqueous species and solid phases in the FMT database are provided as the standard chemical potential (u[0]/RT) of each aqueous species or solid phase, which can be used to calculate the log K values for the appropriate dissociation or dissolution reaction.  The log K for a reaction can be calculated from the change in the chemical potentials of the reactants and products (Δu[0]r/RT) using the relationship:

                            logK= -Δur0/RT2.30259

Calculation of the log K values for the aqueous species and solids in data0.fmt.R0 from the standard chemical potential data in the FMT_050405.CHEMDAT database was described by Ismail et al. (2009) and documented in the spreadsheet DBConversion050405_Rev3.xls.  The log K values in data0.fmt.R0 were verified by independently performing calculations of the log K values from the standard chemical potential data for the AmEDTA[-] and Th(OH)4(aq) dissociation reactions and the anhydrite [CaSO4(s)], ThO2(am), Am(OH)3(s) and hydromagnesite5425 [Mg5(CO3)4(OH2)::4H2O(s)] dissolution reactions.

Some of the standard chemical potential data in FMT_050405.CHEMDAT were originally calculated from log K values in the literature.  As an additional verification of the conversion of the standard chemical potential data in FMT_050405.CHEMDAT to log K data in data0.fmt.R0, the log K values in the original references were compared to the log K data in the EQ3/6 database for the aqueous species AmCO3[+] (Fanghänel and Kim 1998) and NpO2OH(aq) (Fanghänel et al. 1995) and the solid phase ThO2(am) (Felmy et al. 1991) and found to be equivalent.

FMT_050405.CHEMDAT included an arbitrary standard chemical potential value of 500 for the aqueous species NaOH(aq), HCl(aq), HClO4(aq), PosIon(OH)(aq), HNegIon(aq) and an arbitrary standard chemical potential value of 999.999 for the aqueous species UnuCat#1+, UnuAn#2-, UnuAn#3-, UnuAn#4- and UnuNeu#1(aq).  These data were not included in the data0.fmt.R0 database.  In addition, the FMT_050405.CHEMDAT database incorporated data for uranium and plutonium aqueous and solid species.  Because WIPP PA calculations do not include direct calculations of uranium and plutonium reactions, these data were not included in data0.fmt.R0 (Ismail et al. 2009).  Database entries for CO2("solid",DISABLED), H[+](solid) and OH[-]/H2O(solid) are also present in FMT_050405.CHEMDAT with standard chemical potential values of zero; these phases were not included in data0.fmt.R0.  These omissions were appropriate, as these data are not needed for the EQ3/6 verification calculations or WIPP PA calculations.

The solid phases in FMT_050405.CHEMDAT were compared to the solid phases in data0.fmt.R0.  With the exception of the omissions noted above, all solid phases were transferred from the FMT database to the EQ3/6 database.  The americium(III) aqueous species in the FMT database were compared to the americium(III) species in the EQ3/6 database and it was confirmed that all americium(III) species were transferred.

Data0.fmt.R0 does not contain information regarding the sources of the log K data for aqueous species unless they are auxiliary basis species.  The database also does not provide log K data source information for the solid phases.  The data sources should be added to the database, citing either the references that provided the log K data or the sources of the standard chemical potentials used to calculate the log K values.
 Gas Phase Data

FMT_050405.CHEMDAT does not contain gas phase data.  The gas phases included in the data0.fmt.R0 database include methane [CH4(g)], carbon dioxide [CO2(g)], hydrogen [H2(g)], water [H2O(g)] and oxygen [O2(g)].  Data provided for the gas phases include elemental composition, the gas dissolution reaction in terms of the strict basis species and the corresponding log K values.  The source of the log K data in data0.fmt.R0 for gaseous methane, hydrogen and water was data0.ymp.R5.  The data0.ymp.R5 database provides the original data sources for the gas phases and these sources should be documented in the data0.fmt.R0 database.

Gaseous oxygen [O2(g)] is a strict basis species and its dissolution reaction is provided by the data for O2(aq), which is an auxiliary basis species.  Consequently, the data block for O2(g) in the gas phase section of the database simply equates O2(g) with itself.  The gaseous carbon dioxide log K at 25°C was obtained using a standard chemical potential of -159.092 for gaseous carbon dioxide that was hard-coded in FMT Version 2.4.  Calculation of the log K value for the gaseous carbon dioxide reaction in the database was verified using this standard chemical potential.

 Pitzer Parameters

The Pitzer parameters used to calculate activity coefficients in high-ionic-strength solutions and included in the FMT_050405.CHEMDAT and data0.fmt.R0 databases are:

 The A[φ] Debye-Hückel parameter
 Single electrolyte parameter values for cation-anion (ca) pairs: α1, α2, β[0], β[1], β[2] and C[φ]
 Common-ion two electrolyte parameter values for cation-cation (cc') and anion-anion (aa') pairs: ϴcc' and ϴaa'
 Common-ion two-electrolyte parameter values for each cation-cation-anion (cc'a) and anion-anion-cation triplet (aa'c): Ψcc'a and Ψaa'c
 Neutral ion parameter values for each neutral ion-cation (nc) and neutral ion-anion (na) pair: λnc and λna 
 Neutral ion-cation-anion (nca) interaction parameters for each triplet: ξnca

Non-zero Pitzer parameters were extracted from FMT_050405.CHEMDAT (Ismail et al. 2009).  EQ3/6 uses a temperature-dependent model for Pitzer parameters, but all WIPP PA calculations are carried out assuming a temperature of 25°C.  Consequently, only the parameters required for calculating Pitzer parameters at 25°C are nonzero in the data0.fmt.R0 database.

 Debye-Hückel parameter A[φ]

FMT uses a hard-coded Debye-Hückel A[φ] value of 0.39, which is the value cited in Harvie et al. (1984).  However, Plummer et al. (1988) determined that the A[φ] value used by Harvie et al. (1984) was actually 0.392, which is the correct value.  Because the value of 0.39 is used in the FMT calculations, A[φ] was set equal to 0.39 in data0.fmt.R0 to maintain consistency (SNL 2011), but this value should be corrected to 0.392 after the EQ3/6 code validation and verification effort is complete.

 Single Electrolyte Parameters

The single electrolyte Pitzer parameters for cation-anion pairs include α1, α2, β[0], β[1], β[2] and C[φ].  The only conversion of the parameters from FMT_050405.CHEMDAT was for α1 and α2, which are used in the equation to calculate the term B[φ]MX for the electrolyte MX (Pitzer 1991):

     B[φ]MX = β[0]MX + β[1]MX exp(-α1I[1/2]) + β[2]MX exp(-α2I[1/2])

In the FMT database, the entries for the single electrolyte parameters include a code for the α1 and α2 values (Table A-3).

           Table A-3.	FMT Database Codes for α1 and α2 Parameters
                             (Babb and Novak 1997)
                                     Code
                            α1 (kg[1/2]/mol[1/2])
                            α2 (kg[1/2]/mol[1/2])
                                  Electrolyte
                                       1
                                      2.0
                                     12.0
                  1:1, 1:2, 1:3, 1:4, 1:5, 1:6, 2:1, 3:1, 4:1
                                       2
                                      1.4
                                     12.0
                                      2:2
                                       3
                                      1.4
                                     50.0
                                 2:3, 2:5, 3:2


Transfer of the single-electrolyte parameters for the selected "Code 1" electrolytes Na[+]  -  Cl[-], K[+]  -  NpO2CO3[-], Na[+]  -  B4O5(OH)4[2][-], Ca[2+]  -  OH[-] and Ca[2+]  -  HCO3[-] was verified by comparing the α1, α2, β[0], β[1], β[2] and C[φ] values in FMT_050405.CHEMDAT to the values in data0.fmt.R0.  As noted in FMT_050405.CHEMDAT and by SNL (2011), Harvie et al. (1984) provides a Na[+]  -  Cl[-] β[1] value of 0.2644, which is included in the FMT database.  The correct value appears to be 0.2664 (Pitzer 1991, SNL 2011), but the β[1] value of 0.2644 was retained in data0.fmt.R0 to maintain consistency with FMT_050405.CHEMDAT.

For all 2:2 electrolytes, the α1 and α2 parameters in data0.fmt.R0 were examined and found to be correct (Table A-3).  Ca[2+]  -  SO4[2-] and Mg[2+]  -  SO4[2-] are the only 2:2 electrolytes in the FMT database with nonzero β[0], β[1], β[2] or C[φ] parameters.  Transfer of the single-electrolyte parameters for the electrolytes Ca[2+]  -  SO4[2-] and Mg[2+]  -  SO4[2-] was verified by comparing the parameter values in the FMT and EQ3/6 databases.

For electrolytes that do not include a univalent cation or anion and that include a cation or anion with a valence of 3 or higher (Code 3), the α1 and α2 parameters in data0.fmt.R0 were verified by comparison of data in data0.fmt.R0 with the values listed in Table A-3.  Transfer of the single-electrolyte parameters for randomly selected "Code 3" electrolytes Mg[2+]  -  NpO2(CO3)3[5-], Am[3+]  -  SO4[2-] and Th[4+]  -  SO4[2-] was verified by comparing the parameter values in the FMT and EQ3/6 databases.

 Common-Ion Two-Electrolyte Parameters

The common-ion two-electrolyte Pitzer parameters for cation-cation and anion-anion pairs are ϴcc' and ϴaa', respectively.  The ϴcc' and ϴaa' values in data0.fmt.R0 were verified by comparing them to the values in FMT_050405.CHEMDAT.  During the verification process, it was noted that reference for the Cl[-]  -  NpO2(CO3)2[3-] ϴaa' (Fanghänel et al. 1995) is missing from data0.fmt.R0.  This source information should be added to the Cl[-]  -  NpO2(CO3)2[3-] ϴaa' data block.

The common-ion two-electrolyte Pitzer parameters for cation-cation-anion and anion-anion-cation triplets are Ψcc'a and Ψaa'c, respectively.  Transfer of these parameters from FMT_050405.CHEMDAT to data0.fmt.R0 was checked for the randomly selected triplets Na[+]  -  K[+]  -  Cl[-], Na[+]  -  H[+]  -  HSO4[-], Ca[2+]  - Mg[2+]  -  SO4[2-], Cl[-]  -  SO4[2-]  -  Mg[2+], Cl[-]  -  B(OH)4[-]  -  Na[+], SO4[2-]  -  OH[-]  -  Na[+] and HCO3[-]  -  ClO4[-]  - Na[+].  Comparison of these Ψcc'a and Ψaa'c values in the two databases indicate that the data were transferred correctly.

 Neutral Ion Parameters

The Pitzer parameters for neutral ion-cation pairs, neutral ion-anion pairs and neutral ion-neutral ion pairs are λnc, λna and λnn, respectively.  Transfer of the λnc and λna parameters from FMT_050405.CHEMDAT to data0.fmt.R0 was verified by comparing λnc and λna values for the randomly selected neutral ion pairs CO2(aq)  -  Ca[2+], B(OH)3(aq)  -  Na[+], Am(OH)3(aq)  -  K[+], CO2(aq)  -  SO4[2-], H3PO4(aq)  -  H2PO4[2-] and NpO2Acetate(aq)  -  Cl[-].  During the review of the λnc parameters, it was noted that the cited source of λnc for B(OH)3(aq)  -  Na[+] is Harvie et al. (1984).  However, the source citation should be changed to the correct citation, which is Felmy and Weare (1986).

Because FMT_050405.CHEMDAT does not include data for O2(aq), the λnc and λna parameters for O2(aq) were obtained from data0.ypf.R1.  These parameters were verified by comparing the values in data0.fmt.R0 and data0.ypf.R1.  Within data0.fmt.R0, the original source of the data (Clegg and Brimblecombe 1990) is also cited.  Pitzer data were not included for H2(aq) because they were not available in data0.ymp.R5.  No λnn parameter values are listed in FMT_050405.CHEMDAT, and λnn values are consequently not included in data0.fmt.R0.

The neutral ion-cation-anion interaction Pitzer parameter for each triplet is ξnca.  The only ξnca data in FMT_050405.CHEMDAT are for the triplets B(OH)3(aq)  -  Na[+]  -  SO4[2-] and B(OH)3(aq)  -  Na[+]  -  SO4[2-].  Comparison of these ξnca values indicated the data were accurately transferred from FMT_050405.CHEMDAT to data0.fmt.R0.  Because FMT_050405.CHEMDAT does not include data for O2(aq), the ξnca parameters for O2(aq)  -  cation  -  anion triplets were obtained from data0.ypf.R1.  These parameters were verified by comparing the values in data0.fmt.R0 and data0.ypf.R1.  Within data0.fmt.R0, the original source of the data (Clegg and Brimblecombe 1990) is also cited.

 Summary of Database Review Findings

The atomic weight data in the data0.fmt.R0 database are accurate and complete and the aqueous speciation, solid phase solubility and Pitzer parameter data in FMT_050405.CHEMDAT were accurately and completely incorporated in the data0.fmt.R0 database.  However, the database documentation should be corrected and more complete source information should be provided:

 The "99lid" reference for the atomic weight data should be added to the data0.fmt.R0 reference list
 The original source (Shock et al. 1989) of the log K data for the auxiliary basis species O2(aq) and H2(aq) should be included in the data source citations in data0.fmt.R0
 The source citation for the NH3(aq) log K value should be changed from "DBConversion050405.xls" to "DBConversion050405_Rev3.xls" and the original sources of the data used to calculate the log K value (Harvie et al. 1984, Shock and Helgeson 1988, Shock et al. 1989) should be added to the data source citation
 The original source of the perchlorate log K data should be added to the data source citation
 The cited source of the log K data for gaseous methane, hydrogen and water in the data0.fmt.R0 database was data0.ymp.R5; the original sources of the gas phase data should be documented within data0.fmt.R0
 For all aqueous species other than the strict basis species and auxiliary basis species, the log K data sources within data0.fmt.R0 should be provided
 The log K data sources for the solid phases should be provided within data0.fmt.R0
 A reference to Fanghänel et al. (1995) in the Cl[-]  -  NpO2(CO3)2[3-] ϴaa' data block should be provided within data0.fmt.R0
 The source of λnc for B(OH)3(aq)  -  Na[+] in data0.fmt.R0 is incorrectly listed as Harvie et al. (1984); this citation should be corrected to Felmy and Weare (1986)

SNL (2011) identified two parameters in the data0.fmt.R0 database that should be corrected: the A[φ] Debye-Hückel parameter should be changed from 0.30 to 0.392 and the Na[+]  -  Cl[-] β[1] value of 0.2644 should be changed to 0.2664.  Database version data0.fmt.R1 is the same as database version data0.fmt.R0, except these two parameters have been corrected.  This corrected database should be used in future WIPP EQ3/6 calculations after the corrections and additions noted above for data0.fmt.R0 are implemented.




