This international standard was developed in accordance with internationally recognized principles on standardization established in the Decision on Principles

for the
Development of International Standards, Guides and Recommendations issued by the World Trade Organization Technical Barriers to Trade (TBT) Committee.

Designation: C1174 − 17

Standard Practice for

Evaluation of the Long-Term Behavior of Materials Used in
Engineered Barrier Systems (EBS) for Geological Disposal
of High-Level Radioactive Waste1
This standard is issued under the fixed designation C1174; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (´) indicates an editorial change since the last revision or reapproval.

1. Scope considerations, expert judgments, and interpretations of data

1.1 This practice addresses how various test methods and obtained from tests and analyses of appropriate analogs.
data analyses can be used to develop models for the evaluation 1.3.3 For the purpose of this practice, tests are categorized
of the long-term alteration behavior of materials used in according to the information they provide and how it is used
engineered barrier system (EBS) for the disposal of spent for model development, support, and use. These tests may
nuclear fuel (SNF) and other high-level nuclear waste in a include but are not limited to: accelerated tests, attribute tests,
geologic repository. The alteration behavior of waste forms and characterization tests, confirmation tests, and service condition
EBS materials is important because it affects the retention of tests.
radionuclides within the disposal system either directly, as in 1.4 This standard does not purport to address all of the
the case of waste forms in which the radionuclides are initially safety concerns, if any, associated with its use. It is the
immobilized, or indirectly, as in the case of EBS containment responsibility of the user of this standard to establish appro-
materials that restrict the ingress of groundwater or the egress priate safety and health practices and determine the applica-
of radionuclides that are released as the waste forms degrade. bility of regulatory requirements prior to use.
1.5 This international standard was developed in accor-
1.2 The purpose of this practice is to provide a
dance with internationally recognized principles on standard-
scientifically-based strategy for developing models that can be
ization established in the Decision on Principles for the
used to estimate material alteration behavior after a repository
Development of International Standards, Guides and Recom-
is permanently closed (that is, the post-closure period) because
mendations issued by the World Trade Organization Technical
the timescales involved with geological disposal preclude
Barriers to Trade (TBT) Committee.
direct validation of predictions.
1.3 This practice also addresses uncertainties in materials 2. Referenced Documents
behavior models and the impact on the confidence in the EBS 2.1 ASTM Standards:2
design criteria, the scientific bases of alteration models, and C859 Terminology Relating to Nuclear Materials
repository performance assessments using those models. This C1285 Test Methods for Determining Chemical Durability
includes the identification and use of conservative assumptions of Nuclear, Hazardous, and Mixed Waste Glasses and
to address uncertainty in the long-term performance of mate- Multiphase Glass Ceramics: The Product Consistency Test
rials. (PCT)
1.3.1 Steps involved in evaluating the performance of waste C1682 Guide for Characterization of Spent Nuclear Fuel in
forms and EBS materials include problem definition, labora- Support of Interim Storage, Transportation and Geologic
tory and field testing, modeling of individual and coupled Repository Disposal
processes, and model confirmation. E177 Practice for Use of the Terms Precision and Bias in
1.3.2 The estimates of waste form and EBS material perfor- ASTM Test Methods
mance are based on models derived from theoretical E178 Practice for Dealing With Outlying Observations
E583 Practice for Systematizing the Development of
1
This practice is under the jurisdiction of ASTM Committee C26 on Nuclear
Fuel Cycle and is the direct responsibility of Subcommittee C26.13 on Spent Fuel
2
and High Level Waste. For referenced ASTM standards, visit the ASTM website, www.astm.org, or
Current edition approved July 1, 2017. Published August 2017. Originally contact ASTM Customer Service at service@astm.org. For Annual Book of ASTM
approved in 1991. Last previous edition approved in 2013 as C1174 – 07 (2013). Standards volume information, refer to the standard’s Document Summary page on
DOI: 10.1520/C1174-17. the ASTM website.

Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States

1
 C1174 − 17
(ASTM) Voluntary Consensus Standards for the Solution accepted in dictionaries of the English language, except for
of Nuclear and Other Complex Problems (Withdrawn those terms defined below for the specific usage of this
1996)3 practice.
2.2 ANSI Standard:4 3.2 Regulatory and Other Published Definitions—
ANSI/ASME NQA-1 Quality Assurance Program Require- Definitions of the particular terms below are generally consis-
ments for Nuclear Facility Applications tent with the usage of these terms in the context of geological
2.3 U.S. Government Documents:5 disposal of radioactive materials. If precise regulatory defini-
tions are needed, the user should consult the appropriate
NOTE 1—The U.S. government documents listed in 2.3 and referenced
in this practice are only included as examples of local regulations that, governing reference.
depending on the location of the disposal site, may or may not be 3.2.1 backfill—the material used to refill excavated portions
appropriate. Users of this practice should adhere to the regulatory of a repository after waste has been emplaced.
documents and regulations applicable in the licensing location. The
references listed below are explicit examples of local regulations. 3.2.2 buffer—any substance placed around a waste package
Code of Federal Regulations, Title 10, Part 63, Disposal of in a disposal facility to serve as a barrier to restrict the access
High-Level Radioactive Wastes in a Geologic Repository of groundwater to the waste package; and to reduce by sorption
at Yucca Mountain, Nevada, U.S. Nuclear Regulatory and precipitation the rate of eventual migration of radionu-
Commission, latest revision clides from the waste.
Public Law 97-425, Nuclear Waste Policy Act of 1982, as 3.2.3 data—information developed as a result of scientific
amended investigation activities, including information acquired in field
NUREG–0856, Final Technical Position on Documentation or laboratory tests, extracted from reference sources, and the
of Computer Codes for High-Level Waste Management results of reduction, manipulation, or interpretation activities
(1983) conducted to prepare it for use as input in analyses, models, or
calculations used in performance assessment, integrated safety
2.4 International Documents:
analyses, the design process, performance confirmation, and
SKI Report 99:2 Regulatory Perspectives on Model Valida-
other similar activities and evaluations.
tion in High-Level Radioactive Waste Programs: A Joint
NRC/SKI White Paper, Swedish Nuclear Power 3.2.4 disposal—in high-level radioactive waste
Inspectorate, March 19996 management, the emplacement in a geologic repository of
IAEA SSR-5 Disposal of Radioactive Waste – Specific high-level radioactive waste, spent nuclear fuel, or other highly
Safety Requirements, International Atomic Energy radioactive material with no foreseeable intent of recovery,
Agency (IAEA), Vienna, Austria, 20116 whether or not such emplacement permits the recovery of such
IAEA GSG-3 The Safety Case and Safety Assessment for waste.
the Predisposal Management of Radioactive Waste, Inter- 3.2.5 engineered barrier system (EBS)—the man-made, en-
national Atomic Energy Agency (IAEA), Vienna, Austria gineered materials placed within a repository (for example,
20136 waste forms, waste packages, waste canisters, backfill, buffer
SSMFS 2008:37 Swedish Radiation Safety Authority Regu- materials) that are designed to prevent or inhibit migration of
latory Code – General Advice, Swedish Radiation Safety radioactive material from the repository.
Authority, Stockholm, January 30, 20097
3.2.6 geologic repository—in high-level radioactive waste
Finland Government Decree (736/2008) on the Safety of
management, a system which is used for, or may be used for,
Disposal of Nuclear Waste, Radiation and Nuclear Safety
the disposal of radioactive wastes in excavated geologic media.
Authority in Finland (STUK) Helsinki, November 27,
3.2.6.1 Discussion—A geologic repository includes the geo-
20088
logic repository operations area, and the portion of the geologic
3. Terminology setting that provides isolation of the radioactive waste.
3.2.7 high-level radioactive waste (HLW)—generally com-
3.1 Definitions9—Definitions used in this practice are as
posed of highly radioactive materials produced as a byproduct
currently existing in Terminology C859, or as commonly
of the reactions that occur inside nuclear reactors that are
disposed of in a deep geologic repository, such as spent nuclear
3
fuel, and wastes resulting from the reprocessing of spent
The last approved version of this historical standard is referenced on
www.astm.org. nuclear fuel.
4
Available from American National Standards Institute (ANSI), 25 W. 43rd St., 3.2.8 risk-informed—refers to an approach that uses the
4th Floor, New York, NY 10036, http://www.ansi.org.
5
Available from U.S. Government Printing Office, Superintendent of results and findings of risk or performance assessments to
Documents, 732 N. Capitol St., NW, Washington, DC 20401-0001, http:// focus attention on those attributes of a geologic repository
www.access.gpo.gov. commensurate with their importance to safety.
6
Available from International Atomic Energy Agency (IAEA), Vienna Interna-
tional Centre, PO Box 100, A-1400 Vienna, Austria, www.iaea.org. 3.2.9 scientific investigation—any research, experiment,
7
Available from Swedish Radiation Safety Authority (SSMFS), Solna Strandvag test, study, or activity that is performed for the purpose of
96, 171 16 Stockholm, www.stralsakerhetsmyndigheten.se.
8
investigating the material aspects of a geologic repository,
Available from Finlex, www.finlex.fi/en/.
9
See Compilation of ASTM Standard Definitions, available from ASTM including the investigations that support design of the facilities,
Headquarters, 100 Barr Harbor Drive, West Conshohocken, PA 19428. such as EBS post-closure performance models.

2
 C1174 − 17
3.2.10 technical information—information available from environmental variables on material changes (alteration) over
drawings, specifications, calculations, analyses, reactor opera- time, and develop model parameter values.
tional records, fabrication and construction records, other 3.3.11 confirmation test—for the prediction of long-term
design basis documents, regulatory or program requirements behavior of materials, a test for which results are not used in
documents, or consensus codes and standards that describe the initial development of a model or the determination of
physical, performance, operational, or nuclear characteristics parameter values for a model but are used for comparison with
or requirements. predictions of that model for model validation.
3.2.11 waste form—the radioactive waste in its physical and 3.3.12 degradation—any change in a material that adversely
chemical form after treatment or conditioning, or both, (result- affects the ability of that material to perform its intended
ing in a solid product) prior to packaging. function; adverse alteration.
3.2.12 waste package—the waste form and any containers, 3.3.13 empirical model—a model representing observations
shielding, packing, and other absorbent materials immediately or data from experiments without regard to mechanism or
surrounding an individual waste container. theory. An empirical model may be developed by representing
3.3 Definitions of Terms Specific to This Standard: experimental data through regression analysis or may be
3.3.1 The following definitions are defined only for the developed to bound all the observed data.
usage in this practice, and for the explanation of the analyses 3.3.14 extrapolation—the act of estimating long-term mate-
contained herein. rial behavior beyond the range of data collected based on trend
3.3.2 accelerated test—for the prediction of long-term be- determined by empirical observation.
havior of materials, a test that results in an increase either in the 3.3.15 in situ test—tests conducted within a geological
rate of an alteration process or in the extent of reaction progress environment representing a potential repository. A special
when compared with expected service conditions. underground laboratory, called an underground research labo-
3.3.2.1 Discussion—Changes in the expected alteration ratory (URL), may be built for in situ testing or tests may be
mechanism(s) caused by the accelerated test conditions, if any, carried out in an actual repository excavation. In situ tests can
must be accounted for in the use of the accelerated test data. be used to measure the full range of initial repository environ-
3.3.3 alteration—a measurable or visible change in a mate- mental properties and material interactions and under natural
rial affecting its chemical, physical, or radiological properties. conditions.
3.3.4 alteration mechanism—the series of fundamental 3.3.16 mechanistic model—model derived using accepted
chemical or physical processes by which alteration occurs. fundamental laws governing the behavior of matter and energy
to represent an alteration process (or processes).
3.3.5 alteration mode—for the prediction of long-term be-
havior of materials, a particular form of alteration, for 3.3.17 model—a representation of a system or phenomenon,
example, general corrosion, localized corrosion. based on a set of hypotheses (assumptions, data,
simplifications, and idealizations) that describe the system or
3.3.6 analog—for the prediction of long-term behavior of explain the phenomenon, often expressed mathematically.
materials, a material, process, or system whose composition
3.3.18 model validation—the process through which model
and environmental history are sufficiently similar to those
calculations and results are compared with independent mea-
anticipated for the materials, processes, or systems of interest
surements or analyses of the modelled property to provide
to permit use of insight gained regarding its condition or
confidence that a model adequately represents the alteration
behavior to be applied to the material, process, or system of
behavior of waste package/EBS materials under particular sets
interest.
of credible environmental conditions. This provides confidence
3.3.7 attribute test—for the prediction of long-term behav- in the capability of the model to estimate alteration behavior
ior of materials, a test conducted to provide material property under conditions or durations that have not been tested directly.
data that are required as input to behavior models, but are not 3.3.18.1 Discussion—Modelling the behavior of an engi-
themselves responses to the environment, such as density, neered system in a geological disposal facility involves tem-
thermal conductivity, mechanical properties, radionuclide con- poral scales and spatial scales for which no comparisons with
tent of waste forms, and so forth. system level tests are possible: models cannot be ‘validated’ for
3.3.8 behavior—the response of a material to the environ- that which cannot be observed. ‘Model validation’ in these
ment in which it is placed. circumstances implies showing that there is a basis for confi-
dence in the model(s) by means of detailed external reviews
3.3.9 bounding model—for the prediction of long-term be- and comparisons with appropriate field and laboratory tests,
havior of materials, a model that yields values for dependent and comparisons with observations of tests and of analogous
variables or effects that are expected to be either always greater materials, conditions and geologies at the process level. Al-
than or always less than those expected for the variables or though the term validation has been used in a geological
effects being bounded. disposal context, the term “validation” has typically been
3.3.10 characterization test—for the prediction of long-term qualified regarding the limitations of its use in the context of
behavior of materials, a test conducted to establish alteration geologic disposal. Thus, the term ‘validation’ is used sparingly
mechanisms for important processes, measure the effects of in this practice and when used is referring to the activities taken

3
 C1174 − 17
to provide support for and confidence in models used for material properties and alteration processes are likely to be
estimating the performance of materials for geologic disposal important under environmental conditions expected during the
applications. Section 21 provides further discussion on model performance period.
validation (support for and confidence in models).
4. Summary of Practice
3.3.19 predict—estimate the future behavior of a material by
using a model. 4.1 This practice covers the general approach for proceed-
ing from the statement of a problem in estimating the long-term
3.3.20 semi-empirical model—a model based partially on a
behavior of materials, through the development, support, and
mechanistic understanding of an alteration process (or pro-
confirmation of appropriate models, to formulation of the
cesses) and partially on empirical representations of observa-
material performance models. Fig. 1 depicts the various steps
tions using data from experiments.
in developing a model through to confirmation of the models
3.3.21 service condition test—a test that is conducted under during operations and the types of testing that could be used to
conditions in which the values of the independent variables are support model development. This general depiction of model
within the range expected for the actual service environment. development and testing is used to provide an overall perspec-
3.3.22 service condition tests—for the prediction of long- tive for the contents and discussion presented in this practice
term behavior of materials, a test conducted to determine what and is not intended to be applied in an overly restrictive

FIG. 1 Model Development Steps and Testing Support

4
 C1174 − 17
manner. For example, certain tests (for example, service disposal, which means that the siting, design, construction, and
conditions tests) are depicted as supporting model formulation; operation of the repository and appurtenant system components
however, this should not be interpreted that these types of test should be carried out so as to prevent, limit, and delay releases
would also not be able to provide support for other steps in from both engineered and geological barriers as far as is
model development (for example, model support and confi- reasonably possible.
dence). The figure is intended to correlate the types of tests and 5.1.4 The Regulatory Authority in Finland identified the
steps of model development in a general sense. Clearly, some need to support the safety assessment stating that the input data
tests may assist multiple modeling needs and purposes. The and models utilized in the safety case shall be based on
final step in model development (that is, long-term estimates of high-quality research data and expert judgement. Data and
material performance) is correlated to a performance confir- models shall be validated as far as possible and correspond to
mation program that is expected to be implemented during the the conditions likely to prevail at the disposal site during the
operational period and, at least in part, allow for monitoring of assessment period.
the actual materials in the repository environment (for 5.1.5 The Office of Nuclear Regulation in the United
example, waste packages with high-level waste emplaced in Kingdom will regulate an operating geological repository
the repository drifts). The double arrows in Fig. 1 are used to under the Nuclear Installations Act through application of the
represent the iterative nature of testing and model develop- Safety Assessment Principles developed for all nuclear facili-
ment. Although the steps in model development process can ties and the post-closure disposal period will be regulated
also be iterative, the vertical arrows in Fig. 1 are used to under the Radioactive Substances Act by the Environmental
represent the progress of model development to its final step Agency. The two regulators have a Memorandum of Under-
(estimating performance of the materials). Fig. 2 provides a standing outlining how the regulators work together
more detailed depiction of the iterative nature and model (onr.org.uk/wastemanage/position-statement.pdf).
development and testing.
5.2 This practice aids in defining acceptable methods for
making useful estimations of long-term behavior of materials
5. Significance and Use
from such sources as test data, scientific theory, and analogs.
5.1 This practice supports the development of material
5.3 This practice recognizes that technical information and
behavior models that can be used to estimate performance of
test data regarding the actual behavior of EBS materials will by
the EBS materials during the post-closure period of a high-
necessity be based on test durations that are short relative to the
level nuclear waste repository for times much longer than can
time periods required for geologic disposal (for example,
be tested directly. This practice is intended for modeling the
thousands of years and longer). In addition to use in formulat-
degradation behaviors of materials proposed for use in an EBS
ing acceptable long-term performance models data from short-
designed to contain radionuclides over tens of thousands of
term tests are used to support the EBS design and selection of
years and more. There is both national and international
materials. For example, low confidence in a degradation model
recognition of the importance of the use and long-term
for one material may justify the selection of alternative EBS
performance of engineered materials in geologic repository
barrier materials that can be modelled with higher confidence.
design. Use of the models developed following the approaches
It is expected that the data and model will reflect the intended
described in this practice is intended to address established
application of establishing design criteria, comparison of
regulations, such as:
performance assessment results with safety limits, etc. See
5.1.1 U.S. Public Law 97–425, the Nuclear Waste Policy
Section 21 for further discussion on model support and
Act of 1982, provides for the deep geologic disposal of
confidence.
high-level radioactive waste through a system of multiple
barriers. These barriers include engineered barriers designed to 5.4 The EBS environment of interest is that defined by the
prevent the migration of radionuclides out of the engineered natural conditions (for example, minerals, moisture, biota, and
system, and the geologic host medium that provides an mechanical stresses); changes that occur over time, during
additional transport barrier between the engineered system and repository construction and operation, and as a consequence of
biosphere. The regulations of the U.S. Nuclear Regulatory radionuclide decay, namely, radiation, radiation-induced
Commission for geologic disposal require a performance damage, heating, and radiolytic effects on the solution chem-
confirmation program to provide data through tests and istry; and changes that may occur over the post-closure period.
analyses, where practicable, that demonstrate engineered sys- Environmental conditions associated with disruptive events
tems and components that are designed or assumed to act as (for example, mechanical stress from seismic events) and
barriers after permanent closure are functioning as intended processes (for example, changes in water chemistry) should
and anticipated. also be considered.
5.1.2 IAEA Safety Requirements specify that engineered
barriers shall be designed and the host environment shall be 6. General Procedure
selected to provide containment of the radionuclides associated 6.1 The major elements in the approach to develop models
with the wastes. for estimating the long-term behavior of EBS materials are
5.1.3 The Swedish Regulatory Authority has provided gen- problem definition, testing, modeling, performance estimate,
eral advice to the repository developer that the application of and confirmation. Fig. 2 is a flow chart showing the logical
best available technique be followed in connection with approach for model development followed in this practice.

5
 C1174 − 17

FIG. 2 Logic for the Development of Models for Estimating the Alteration Behavior of Materials

Although it is not expected that the structure of Fig. 2 will acceptable alteration models, it is likely that the development
apply exactly to every situation, especially as to the starting of models for most materials will contain these major elements.
point and the number and type of iterations necessary to obtain Details on the individual elements are given in Sections 7 – 26.

6
 C1174 − 17
Development of performance models will likely be conducted PROBLEM DEFINITION
under a quality assurance program as discussed in Section 27.
An important aspect of performance models is the uncertainty 7. Scope
of the model, including uncertainties in the form of the model, 7.1 The objective of the problem definition is to identify the
the data used to determine model parameters, and the environ- materials and environments to be assessed and the processes,
mental service conditions to which the model is applied. The interactions, and alteration modes that should be included in
consequences of these uncertainties with regard to the perfor- the models. This information is used to design conceptual
mance of the disposal system are used to determine the models and design tests to develop and evaluate process
uncertainty in the risk. These are discussed in Section 24. models. An extensive list of features, events, and processes
6.2 Identification of Materials: (FEPs) that should be considered has been compiled and
utilized world-wide; however, many of these FEPs lists tend to
6.2.1 The various materials to be evaluated for use in the
be more generic than specific to a particular site or material. A
systems, structures, components, and barriers that are designed
generic FEPs list is a reasonable starting point for developing
and deployed to contain radionuclides within the repository
more site and material specific FEPs that would be expected to
environment must be identified. A risk-informed approach to
address the specific materials and site conditions being inves-
repository performance assessment can be used to identify the
tigated.
behavior characteristics of those materials that may substan-
tially contribute to risk by affecting the release of radionuclides 7.2 In this practice, methods are recommended for the
from the repository over the post-closure period. Performance development of performance models for long-term alteration of
assessments can analyze the sensitivity to specific materials EBS materials that are proposed for use in the geologic
and alteration processes and disruptive events (for example, disposal of high-level radioactive wastes. This practice recom-
seismic activity) to identify the attributes of particular EBS mends a methodology for assessments of performance of
materials that are most important for limiting the release of materials proposed for use in systems designed to function
radionuclides over the long time periods of geologic disposal. either for containment or control of release rates of radionu-
It is the long-term behavior of these risk-significant materials clides.
that is the subject of this procedure. 7.3 Problem definition includes identifying factors that are
6.2.2 Modeling the alteration behaviors of EBS materials important in the development of models to support evaluations
having degradation characteristics that are determined to be of long-term behavior of repository materials during the
important to waste isolation needs to be performed with post-closure period. This can be done using literature surveys
sufficient accuracy and precision to determine the useful and other sources of information helpful in characterizing the
lifetimes and expected performance of these materials. All alteration of EBS materials. The key factors include the
relevant degradation processes need to be understood suffi- following:
ciently so that the impact of these materials is not under- 7.3.1 Identification of potential environmental conditions to
estimated and modeling outputs can be used to provide reliable which the material may be exposed,
input to risk-based decision making / optimization. The altera- 7.3.2 Identification of possible EBS design concepts,
tion behaviors of EBS materials having degradation character- 7.3.3 Identification of EBS materials,
istics that are determined to be unimportant to waste isolation 7.3.4 The identity, composition, and condition of the waste
do not need to be modelled with the same accuracy and forms,
precision as those materials deemed to be important to waste 7.3.5 Identification of potential materials alteration modes,
isolation. and
6.3 Identification of Credible Ranges for Environmental 7.3.6 Identification of appropriate natural analog materials.
Conditions: 7.4 This practice outlines a logical approach for estimating
6.3.1 The alteration behavior of a material will depend on the behavior of materials over times that greatly exceed the
the environment in which it is used. The environment within a time over which direct experimental data can be obtained. It
disposal system will be affected by both the natural conditions emphasizes accelerated tests and the use of models that are
and events, the design and materials used in the EBS, and by based on an appropriate mechanistic understanding of the
the alteration of EBS components. For example, the chemistry processes involved in long-term alterations of materials used
of groundwater that contacts the waste forms will be signifi- under repository conditions.
cantly affected by reactions with the natural materials, the
thermal effects of waste emplacement, corrosion of EBS 8. General Considerations
materials, and radiolysis. The anticipated range of repository 8.1 Site Characterization—A potential repository site must
environments throughout the post-closure period should be be investigated with respect to its geologic, hydrologic,
defined and the model developed using test results representing seismic, etc. conditions that could affect the performance of the
this range to the extent practical. repository. For purposes of this practice, site characterization

7
 C1174 − 17
includes the identification of likely impacts of the environmen- 8.4.2 Bounding Conditions—Bounding conditions represent
tal conditions on the behavior of the EBS materials (see 8.5.1, the anticipated extreme credible values of a range of parameter
9.1, and 10.2). or variable values. These furnish necessary input for estimating
8.1.1 Environment—The geologic environment shall be performance limits. However, thorough evaluations of the
evaluated by characterization of the initial environment and alteration mechanisms, all important material attributes, and
mechanical condition and consideration of the effects of time the effects of these attributes on the anticipated alteration
and alteration of EBS and waste form materials on the processes are required to ensure that the calculations represent-
environment. Ranges in the values of such environmental ing bounding conditions do indeed provide performance limits.
conditions as temperature, groundwater chemistry, For example, the pH value that gives the lower limit of the
microbiology, colloid content, and disruptive events (for glass dissolution rate (for example, pH 7) may not be the
example, seismic activity) may be needed to account for extreme value of the range of environmental pH values
changes in the environmental conditions that occur over time. considered (for example, pH 3). Additionally, it is important to
A special underground laboratory, called an underground ensure that the combination of boundary conditions/parameter
research laboratory (URL), may be built to enhance character- values that are considered avoid non-physical or contradictory
ization activities and for in situ testing or tests to be carried out conditions that could lead to unrealistic model results, such as
in a representative repository excavation. large volumes of water being present at temperatures exceed-
ing the local boiling point.
8.2 Conceptual Designs—A general concept for an EBS
design can be initially developed to meet regulatory require- 8.5 Preliminary Testing—A substantial amount of data re-
ments based on current understanding of: (1) the conditions of lated to both the materials of interest, including the waste
a particular site, and (2) the performance of EBS materials forms, and the extant environmental conditions may be avail-
under the site conditions. able before the initiation of tests for model development.
Various preliminary modeling and testing efforts can be con-
8.3 Materials Identification—From the initial concepts and ducted to understand specific aspects of the material/
investigations of a repository site, candidate EBS component environment system and make preliminary evaluations of the
materials are proposed based on the geologic environment and alteration processes. Insight gained from the preliminary tests
the conceptual design. Since these materials serve the function and evaluations can be used to design characterization and
of containment and control of potential radionuclide release accelerated tests for use in the development of the model for
rates, their alteration behavior under the set of conditions long-term performance.
expected in the repository over long time periods must be 8.5.1 Interactions—The process of predicting materials be-
reliably determined and the alteration modes understood. This havior in repositories must involve consideration of interac-
understanding is developed by first reviewing both the avail- tions between materials and environments. For example, inter-
able information regarding the environmental conditions and actions between various materials and the environment may
the effects of the environment on the candidate materials. lead to the formation of reaction products that, in turn, become
8.3.1 Information regarding natural analogs might be avail- part of the environment. Interactions between different mate-
able to provide early guidance for the selection of EBS rials within the EBS may be direct, in the case of materials that
component materials and the long-term alteration of these are in physical contact, or indirect through the groundwater
materials in the repository environment. chemistry. That is, changes in the groundwater due to corrosion
8.3.2 The selection of materials for the EBS could be of one material will affect the corrosion behavior of other
influenced by the support and confidence for degradation rate materials that the groundwater contacts. Of course, it is
models. This approach could lessen the need for hard-to- possible that thermal or mechanical effects on EBS materials
achieve high confidence levels in a degradation model. For could be more important than corrosion processes, which could
example, a container material that exhibits a moderate but increase the significance of seismic events. Characterization
predictable rate of general corrosion, but is not susceptible to tests should be conducted to ensure that the range of environ-
localized corrosion, might be selected for use as a corrosion mental parameters represents the impacts of relevant processes
barrier and the thickness of the wall engineered to provide for and events.
a ‘corrosion allowance.’ 8.6 Literature Survey—Using the proposed materials and
8.4 Ranges of Materials Properties and Environmental estimates of environmental conditions, a literature survey shall
Conditions—Preliminary descriptions of the materials to be be conducted to obtain insight into possible alteration modes
tested shall be used to determine their physical and mechanical and possibly data that can be used in the development of a
properties. Frequently, a range of values will be needed to model. A literature survey must be conducted to identify and
specify parameters used to characterize materials. evaluate the usefulness of any analogs for later testing and
8.4.1 Ranges—A range of parameter values for environmen- evaluation activities.
tal conditions or material properties may be used to account for 8.7 Preliminary Models—For each important alteration
uncertainty. For example, environmental conditions may in- process, preliminary models shall be developed to represent
clude the anticipated temporal and spatial variability, and the and evaluate steps in the process, postulates, and inferences
waste forms may be described by ranges that take into account related to either observed or expected behavior of the materials
differences in properties due to variations in composition in the proposed environments. Preliminary models could use
production history, product usage, process control. conservative approaches that would be used to help focus

8
 C1174 − 17
further model development and data collection in those areas tation processes that convert the glass to phases that are
that are most important to safety. More realistic models (that is, thermodynamically stable. For the alteration mode of glass
less conservative) could evolve as model development and data dissolution, one can describe an alteration mechanism that
collection proceeds. More realistic analyses would provide includes water diffusion into the glass and various reactions
insight into the conditions that may occur and insights into the associated with ion-exchange and hydrolysis. For precipitation
safety margins of bounding assessments. processes, an alteration mechanism for the formation of altera-
8.7.1 Inputs to these models can be estimates of values for tion phases could include precipitation from solution or phase
the independent variables pertinent to environmental condi- transformation of a gel.
tions and alteration processes or values that are obtained from 9.7 Identify Potential Analogs—Identify potential analogs
experiments or other sources. The models are used to estimate for materials, processes, or systems. These may be either
pertinent dependent variables, as for example, dissolution rate natural or man-made.
as a function of time. 9.7.1 Identify the aspect of the analog that can be compared
9. Specific Procedure—Problem Definition (See Fig. 1) with the material or process under consideration. Differences
will likely exist between the compositions of the analog and the
9.1 Define Credible Range of Environmental Conditions— repository material and the environment to which they are
Determine the range of environmental conditions to which the exposed. Evaluations of the significance of the differences may
material will be exposed during (1) the operational period, as be used to support or disqualify use of the analog as a means
appropriate, and (2) after permanent closure (that is, the for providing confidence in the alteration model.
post-closure period). The range should include initial environ-
mental conditions and changes that will occur over time due to TESTING
changes in climate, radiolysis of air and groundwater, corro-
sion of EBS components, etc. The extent of such interactions 10. Scope
may be difficult to quantify initially, but should be noted and 10.1 Model Confidence—The confidence in model results
accounted for in a final model. will depend upon both how well the model represents the
9.1.1 Features, Events, and Processes (FEPs) relevant to alteration mechanism under the in-service conditions (for
degradation and alteration of the EBS components should be example, type or stoichiometry of corrosion product, form of
identified. The FEPs can be used to determine the range of alteration layers, mode of degradation), how well the depen-
environmental conditions (for example, temperature, chemical dencies on environmental variables are represented in the
constituents, and mechanical loads) to help identify the degra- model, and how well the values of environmental variables
dation processes to be evaluated and relevant test conditions. used in the model represent the in-service environmental
9.2 EBS Conceptual Design—Establish the design concepts conditions (for example, temperature, groundwater chemistry,
of the EBS and propose the functional and spatial relationship groundwater quantity).
for the various components. 10.1.1 The ability of the behavior model to provide reliable
9.2.1 If viable options exist in the EBS conceptual design, estimates will be strongly dependent on the accuracy with
activities to address performance issues pertinent to each which the mathematical form of the model represents the
option can be incorporated into subsequent modeling and process kinetics (for example, the degree to which the model is
testing steps to inform future decisions. For example, the based on a mechanistic understanding of the alteration
values of some parameters will differ depending upon whether process), uncertainties in the test data used to derive the
emplacement geometry is vertical or horizontal. parameters and parameter values used in the model, and the
9.3 Identify EBS Materials—Identify the types and intended uncertainties in representations of the actual in-service condi-
uses of all the materials that comprise the EBS components. tions for which the model is applied (see Section 24 on
This would include, for example, identification of weldments Uncertainties).
and the processes and materials with which they are to be 10.1.2 Testing of EBS materials is required to establish the
fabricated. effectiveness of these materials to contain radionuclides in the
repository environment or limit their releases, or both. Tests
9.4 Identify Possible Alteration Modes—Use technical lit- conducted over a comparatively short period, for example, less
erature to help identify possible alteration modes for the than 20 years, will be used to support development of perfor-
materials of interest relevant to the environmental conditions mance models for materials behavior in the repository envi-
for the repository site being evaluated. ronment. The testing program must address the development,
9.5 Identify Variables—Identify the variables regarded to be scientific basis, and confirmation of these models.
important to material behavior in the disposal system, for 10.1.3 Materials testing programs should be designed with
example, the amount of water expected to contact a waste the goal of supporting the validation and verification of
glass. For each independent variable, identify the expected materials behavior models, as well as minimizing uncertainties
range of values. in the test data, the models, and the use of the models in
9.6 Identify Possible Mechanisms for Alteration calculations of long-term behavior in the repository environ-
Processes—For each alteration process, identify possible al- ment.
teration mechanisms to be evaluated by testing and modeling. 10.2 This practice does not address testing required to
For example, glass may be altered by dissolution and precipi- define (or model) the repository design or environment (that is,

9
 C1174 − 17
the groundwater quantity or chemistry, host rock properties, cliff-edge effects just outside the expected parameter ranges,
etc.). The testing concepts described herein do not specifically and (2) demonstrating continuity of mechanisms over the
address the testing of integrated systems within the EBS. It is ranges used in accelerated tests.
expected that the logical approach in this practice can be 13.2 Specific Procedure-Characterization Tests:
applied to integrated systems. 13.2.1 Use literature analyses, analogs, scientific judgment,
10.3 Types of Tests—Testing of EBS materials will be and experience to postulate potential material alteration modes
required for a variety of reasons and thus are expected to and mechanisms.
include a variety of tests, such as: attribute tests, characteriza- 13.2.2 Perform tests to identify alteration mechanisms that
tion tests, confirmation tests, and service condition tests. occur in the repository environment conditions.
13.2.3 Analyze the quantitative and qualitative information
11. Reserved from the characterization tests and identify the alteration
mechanism(s) occurring under the test conditions.
12. Attribute Tests 13.2.4 Identify material and environmental variables affect-
12.1 General—Estimation of the response of materials to ing the alteration rate. Conduct tests using ranges of values to
the repository environment during the post-closure period will determine the kinetic dependencies.
require the specification of the intrinsic properties (“attri- 13.2.5 Integrate the results of characterization tests with the
butes”) of the materials. These properties are not expected to behavior modeling (see Modeling section).
change over time in response to the repository environment.
12.1.1 Examples of material attributes are density, thermal 14. Accelerated Tests
conductivity, chemical composition, radionuclide content, me-
14.1 General—The purpose of an accelerated test is to
chanical properties, etc.
increase the rate of one or more alteration process or the
12.1.2 Attribute tests are designed to provide specific infor-
reaction progress without changing the mechanism(s) of the
mation on test materials necessary for the development of the
alteration process under investigation. Therefore, some knowl-
behavior models when reliable data are not available from the
edge of the mechanism that is operative under in-service
literature. It is expected that most of the required information
conditions is needed for the design of the accelerated test and
concerning barrier materials (for example, steels), spent fuel,
meaningful use of accelerated test data. Processes may be
and high level waste material attributes will be available in the
accelerated by changing various test parameters relative to
literature, but measurements of some properties may be re-
their in-service values, including temperature, material surface
quired.
area, mechanical loads, solution volume or flow rate, initial
12.2 Specific Procedure-Attribute Tests: solute concentrations, humidity, etc. Care should be taken to
12.2.1 Identify the material properties required to apply the ensure, to the extent practical, that the test method and test
model. conditions do not alter the mechanism of the process that is
12.2.2 Examine the literature for materials properties and being accelerated (for example, characterization tests, as dis-
evaluate which properties may be unambiguously determined cussed in Section 13, may be useful in identifying potential
without testing. limitations in accelerated tests).
12.2.3 Perform attribute tests on those properties for which 14.1.1 If the alteration mechanism that is operative in the
unambiguous values could not be determined from the litera- accelerated test differs from that which is operative under the
ture. in-service conditions or changes over a range of accelerating
12.2.4 Compile the values for all properties that may be test conditions, the accelerating test conditions and response
required as input to modeling. must be evaluated to determine if and how the change is related
to the process being accelerated. In many cases, changes in the
13. Characterization Tests process can be detected using trends in the response as the
13.1 General—Characterization tests have the primary accelerating test parameter is varied.
function of providing a mechanistic understanding of the 14.1.1.1 Temperatures higher than the expected service
important processes of material alteration expected in the conditions are often used to accelerate the rate of corrosion of
repository environment and measuring model dependencies a material. The effect of increasing the test temperature can be
and parameter values. These tests are used to establish both the represented using an Arrhenius plot to detect changes in the
suitability and the basic mathematical form representing the effective activation energy, which may indicate a change in
process in the behavior model. mechanism.
13.1.1 Purpose—Characterization tests are designed to 14.1.1.2 Other test results indicate changes in mechanism
identify EBS alteration mechanisms that could occur in a that may or may not impact the process being evaluated.
repository and the dependence of those processes on environ- Consider a series of accelerated tests conducted at different
mental conditions. temperatures in which dissolution of a primary phase resulted
13.1.2 Test conditions may differ significantly from the in formation of corrosion product A at repository-relevant
expected repository conditions, and so it may be necessary to conditions but in formation of corrosion product B at tempera-
investigate the sensitivity of the alteration mechanisms to tures above a critical temperature T°. If the process being
variations in the values of particular test parameters. Extending accelerated is affected differently by formation of corrosion
test parameter ranges could also be useful for: (1) evaluating products A and B, for example, by the release of the soluble

10
 C1174 − 17
species j into solution, the accelerated tests in which B forms in the development of a reliable performance model. In
are not applicable. If which product phase forms is irrelevant to general, the steps given in 14.2 should be followed.
the dissolution rate of the primary phase, the accelerated tests 14.2 A Specific Procedure for Accelerated Testing:
above T° may be applicable. 14.2.1 Define possible alteration mechanisms.
14.1.2 Use—Accelerated tests may be used to: 14.2.2 Identify the alteration process to be accelerated,
(1) Alter the state of a material in a short time to simulate method to accelerate, parameters that can be used for
exposure to repository conditions over long time periods, and acceleration, and alteration indicators (for example, extent of
thereby produce artificially “aged” materials. (This may be corrosion based on weight change).
desirable for determining the characteristics of materials after 14.2.3 Identify the type of test(s) and range of test condi-
long exposures to potential repository conditions or for testing tions to be used in the accelerated test (for example, select
the response of “aged” materials to possible changes in the conditions to isolate the effects of an individual variable).
repository conditions, such as after a large seismic event.), 14.2.4 Perform tests using a set of parameter values ex-
(2) Measure the rates of slow reactions within reasonable pected to increase rate of process relative to service conditions.
laboratory time-scales, 14.2.4.1 Compare the nature and extents of alteration at-
(3) Promote the formation of alteration phases for identi- tained within the series of tests conducted using the range of
fication and characterization, accelerating parameter values and, if relevant, with alteration
(4) Promote the approach to solution saturation, and attained in tests using parameter values that represent service-
(5) Age the solution that contacts the material to represent conditions.
conditions that may occur after long reaction progress. 14.2.4.2 Verify that the variations in process evaluated using
14.1.2.1 An example is the exposure of samples of spent the accelerated tests is relevant and can be related to the
fuel to conditions that accelerate alteration relative to reposi- mechanisms expected to be operative under service conditions.
tory conditions (such as high temperature, high solution Eh, 14.2.5 Identify the range of test conditions to which the
crushing to expose grain boundaries and increase surface area, accelerated behavior applies and compare with the postulations
etc.) to obtain upper limit values for radionuclide release upon in 14.2.2.
exposure to groundwater. The effects of each accelerating 14.2.5.1 Show that the process is relevant to the mecha-
condition on the dissolution rate should be quantified and nisms expected to occur under disposal conditions, taking into
mechanistically described. account anticipated changes in the environment to which the
materials of interest will be exposed.
14.1.3 Synergistic or Competing Effects—Because of the
14.2.5.2 If the alteration mechanisms observed in the accel-
potentially large number of independent variables that affect
erated tests differ from those assumed in the process model,
material alteration (for example, temperature, radiation, me-
reevaluate the relevance of the process model, the test method,
chanical stress, fluid chemistry, and material condition), careful
and the test conditions used to accelerate the process to the
consideration should be given to possible synergistic and
service conditions and return to 14.2.2 to iterate on this
competing effects of the accelerating conditions.
procedure until a satisfactory accelerated test method is devel-
14.1.4 Models—Results of accelerated tests can be used to oped.
develop or support a performance model by demonstrating 14.2.6 Provide results as input to the modelling activity.
conditions under which materials perform well or perform
poorly. They can also be used to demonstrate when an 15. Service Condition Tests
alteration process can be excluded from the model or provide
bounding parameter values. 15.1 General—The purposes of service condition testing are
to determine variables that affect corrosion behavior, identify
14.1.4.1 As an example of excluding a process, a test for
those that must be represented in the alteration model (either
stress corrosion cracking (SCC) of a candidate waste container
explicitly or implicitly by using lumped variables), and estab-
material might establish that the initiation of SCC can only
lish a database for determining the alteration mechanism
occur under temperatures that are higher and aqueous chemis-
operative under repository-relevant conditions. These may
tries that are more aggressive than those that can plausibly
include laboratory tests under conditions simulating disposal
occur in the repository. conditions, lysimeter tests, tests in underground research
14.1.4.2 An example of alteration model parameter mea- laboratories, burial tests, etc.
surement might be a test for general corrosion that is conducted 15.1.1 These tests are used to identify the key aspects of the
at a higher level of anodic polarization than expected to occur materials and the environment that affect the alteration mecha-
in the repository. From the data, best-fit values could be nisms under expected conditions. Observations of the altera-
obtained for making a determination of an bounding corrosion tion mechanisms under service conditions can be used to
current density using a mathematical model for general corro- determine the relevance of accelerated tests results (and the
sion that incorporates passivation and passivation breakdown mechanisms observed therein) for developing alteration mod-
processes. This would provide support for and confidence in els and deriving alteration model parameter values.
using the model for long-term assessments. 15.1.2 Service condition tests should be designed to mea-
14.1.5 Fig. 2 shows the steps involved in the development sure the dependence of material behavior on as many poten-
and performance of accelerated tests. The figure also empha- tially relevant environmental conditions as practical to identify
sizes the necessary connection between testing and modelling important environmental variables.

11
 C1174 − 17
15.1.3 Service condition tests establish the values of key 16.1.3.1 A good use of analogs would be to provide addi-
environmental variables to be used as the reference case for tional confidence in the sensitivity of model results to a range
long-term confirmation testing (see Section 13). of material and environmental conditions. It is unlikely that
15.1.4 Service condition tests may provide data for altera- analogs will be found that are identical in composition and
tion of materials under actual repository test conditions with conditions of exposure to the EBS materials in the repository.
which models can be confirmed, for example, short-term For example, natural uranium minerals might be used as
in-situ tests conducted in underground research laboratories. analogs for the alteration of uranium dioxide spent fuel, but
15.1.5 The configurations of service condition tests are such an analysis should recognize that such minerals did not
likely to be similar to those of the confirmation tests (as evolve in a geochemical environment that included close
described in Section 17) with the primary difference being the proximity to zirconium metal. The sensitivity of test responses
test duration. The duration of a service condition test per- of natural uranium minerals in the presence and absence of
formed for model development may be extended to serve zirconium would indicate their usefulness as analogs.
model confirmation purposes (see Fig. 1). 16.1.4 Characterization of the short-term behavior of analog
materials in laboratory experiments could be used to establish
15.2 Specific Procedures-Service Condition Tests: that the analogs behave similarly in natural and experimental
15.2.1 Select test conditions. “Normal” test conditions may environments. This would support the conclusion that the
be defined in terms of ranges that include maximum, average, relevant mechanisms have been taken into account in the
and minimum values anticipated for each key variable. model. However, any conclusions must give due consideration
15.2.1.1 Conduct sufficient number of tests to measure to survivor bias as well as the representativeness of the
responses spanning the full range of normal conditions for each exposure conditions of the analog to the material under study.
variable.
16.2 Specific Procedure-Section and Testing of Analogs:
15.2.1.2 Compile and evaluate the data obtained for under- 16.2.1 Identify Analogs—Identify natural or man-made ana-
standing of the materials alteration behaviors. Results obtained logs appropriate for the material and alteration mode under
in tests conducted under normal environmental conditions may investigation.
be used as reference values for tests conducted under condi- 16.2.1.1 Search existing literature for potential analogs.
tions outside the normal range to accelerate alteration, under- Include work in other disciplines, such as archaeo-metallurgy,
stand the alteration mechanism, or evaluate the dependence on geology, and history.
key variables. 16.2.1.2 Analyze the degree of similarity and evaluate the
usefulness of the analog in providing information for the
16. Analysis and Testing of Analogs alteration mode of interest.
16.1 General—When estimates of long-term performance 16.2.2 Samples—Obtain multiple samples of the proposed
are made based on models obtained using the results of analog materials, including samples of differing ages and
characterization, accelerated, and service condition tests, con- differing degrees of alteration, if applicable and available.
fidence in the performance estimates over many thousands of 16.2.3 Characterize the site where the analogs were found,
years could be considerably enhanced through the analyses of for example:
natural and man-made analogs. For analog materials to be 16.2.3.1 Dating of site,
useful, reliable information should be available concerning 16.2.3.2 Geology of site and depth of burial,
their age, chemical composition, and exposure history. The 16.2.3.3 Sample storage conditions following retrieval, and
material properties can be determined by using attribute testing 16.2.3.4 Site environment (soil, precipitation, air, etc.).
as described in Section 10, but determination of exposure 16.2.4 Characterize the analogs, including:
conditions, such as solution compositions, contact time, and 16.2.4.1 Photographic documentation of specimens and of
temperature, is outside the scope of this practice. retrieval process,
16.1.1 Choice—Analogs should be chosen with the under- 16.2.4.2 Dating of specimens and time of exposure,
standing that it is likely that no perfectly matching analog will 16.2.4.3 History of specimens and environmental exposure,
be found. For example, no compositional analog to stainless including nature of water contacting material, contact time,
steel is expected that is over 100 years old, iron objects exist, temperature, etc.,
including enriched in nickel, that may have some applicability 16.2.4.4 History of conditions of formation or manufacture,
to selected alteration behaviors. if applicable and available,
16.1.2 The analyses of analogs can be useful in determining 16.2.4.5 Bulk chemical composition analysis of analog,
whether different mechanisms can control alteration over long 16.2.4.6 Surface layer composition analyses (SEM, EDS,
time periods. etc.), and
16.1.3 Use—Natural and man-made analog materials can 16.2.4.7 Structural analyses (microstructure, grain size,
serve as the test specimens for the characterization tests crystallinity, size, shape, color, etc.).
described in Section 13 and the accelerated tests described in 16.2.5 Perform attribute, characterization, accelerated, and
Section 14. The test responses of analogs can provide confi- service-condition tests, as required.
dence in an experimental method for accelerating corrosion 16.2.6 Analyze the data, for example:
behavior and in the models used for particular alteration 16.2.6.1 Estimate the rate of alteration of the analogs,
modes. 16.2.6.2 Determine the mechanism(s) of alteration,

12
 C1174 − 17
16.2.6.3 Compare the data from tests of analogs with data estimate the long-term corrosion behavior of materials based
from tests of the candidate materials or waste forms, and on physical laws, conceptual models, and relatively short-term
16.2.6.4 Use the results of these data analyses in the experimental observations to provide data to derive the model,
development and validation of the models. and insights from natural analogs.
18.2.1 It is expected that development of models and
17. Confirmation Tests generation of test data will be an iterative process. Preliminary
17.1 General—Confirmation tests are designed to produce models could use conservative approaches that would be used
materials alteration data to support application of the alteration to help focus further model development and data collection on
model to the EBS system after the initial formulation and use those aspects determined to have the greatest impact on safety.
of the model developed for demonstrating compliance reposi- More realistic models (generally less conservative) could
tory safety during the post-closure period. Testing (particularly evolve as model development and data collection proceeds.
in-situ testing) should be continued as long as practical and 18.3 General considerations in modeling and specific pro-
necessary during the pre-closure period of the repository but cedures are addressed in this practice.
prior to permanent closure of the repository, to confirm key
aspects of the behavior of the EBS materials used in models for 19. General
estimating the EBS performance during the post-closure pe- 19.1 Function of Modeling—Modeling serves at least two
riod. Also, tests that had begun as service condition tests could functions: demonstration of the self-consistency of data (inter-
be extended to serve the purpose of confirming the estimated polation) and estimation of long-term behavior (extrapolation).
materials alteration behavior.
19.2 Types of Data Used in Modeling—This practice pro-
17.1.1 Use—Confirmation tests, which are to be conducted
vides for the use of several types of information and data in the
prior to permanent closure of the repository, are used to
development and application of models:
provide data showing the alteration model is appropriate for
19.2.1 Characterization test data,
representing material behavior during the post-closure period.
19.2.2 Accelerated test data,
17.1.1.1 They would generally be conducted in-situ (such
19.2.3 Service condition test data,
as, within an exploratory shaft facility at the repository site) or
19.2.4 Analog test data,
under the full suite of conditions expected to be present within
19.2.5 Confirmation test data, and
the repository. Confirmation testing provides further support
19.2.6 Literature information.
for the integrated alteration behavior of materials independent
of the data collected to support license application analyses. 19.3 Types of Models—Quantitative models may range from
purely empirical to purely mechanistic, depending on the
17.2 Specific Procedure-Confirmation Tests:
degree to which the mechanisms of the material alteration
17.2.1 Identify and directly measure repository in-service
processes are explicitly represented in the model.
environmental parameters, such as temperature and groundwa-
19.3.1 Mechanistic Models—In purely mechanistic models,
ter chemistry.
the dependent variables are related to independent variables
17.2.2 Identify the material alteration mode to be
through individual or coupled processes that have been
investigated, the manner of testing, and the behavior model to
identified, are understood, and have scientific bases. The
be confirmed.
relationships are expressed using mathematical representations
17.2.3 Perform tests (in-situ, as appropriate) and observe the
for chemical or physical processes. A purely mechanistic
alteration under repository conditions.
model for a process can be represented mathematically by Eq
17.2.4 Examine material alteration and compare with the
1:
behavioral model results (see Performance Confirmation Sec-
tion 25). If the comparison is not satisfactory, it will be Y 5 F ~ x 1… x n! , (1)
necessary to return to the Modeling section of this practice, as where Y is a dependent variable and xi through xn are all
this is an iterative process. independent variables that affect the value of Y. The expression
17.2.5 Compile confirmation test results and integrate into F(xi) may be comprised of separate terms to represent the
uncertainty analyses of long-term behavior model(s). contributions of different coupled processes. The dependence
MODELING of the response on an individual variable xi is usually deter-
mined by evaluating the results of characterization, service
18. Scope condition, and accelerated tests designed to isolate or highlight
the effect of that variable.
18.1 Modeling may be performed on a risk-informed basis 19.3.1.1 Mechanistic relationships may be identified
to estimate the effects of alteration processes on systems, through first principles and a series of tests (usually
structures, and components that contribute to waste isolation. accelerated, characterization, and service condition tests) to
Modeling may also be performed in support of EBS designs. measure the effects of particular variables (xi) on the test
18.2 A model is used to represent the material alteration response (Y) and attributed to specific alteration processes.
behavior measured by the responses (the dependent variables) Mechanisms can be proposed and evaluated for each specific
in various tests to variables that have been found to be step or process that occurs and then combined into an overall
significant (the independent variables) using mathematical mechanism. The proposed mechanism should identify the roles
expressions. The objective of modeling in this practice is to of all variables that significantly affect the alteration rate to be

13
 C1174 − 17
considered as a purely mechanistic model. In most cases, the representative of the processes expected to have the greatest
values of model parameters are extracted from characterization impact on long-term behavior. Relationships between the
tests conducted specifically for that purpose and verified using dependent and independent variables having the form of Eq 2
other tests in which several variables may affect the material can be inferred by scientific reasoning that describes those
response. For example, if the dissolution rate of a material is steps. This is done by conducting characterization tests to
known to depend on the temperature, pH, and chloride ion measure the effects of important variables and determine the
concentration in solution, tests to determine the effect of forms of the functions f(xi) that minimizes the residual term.
temperature would be conducted at various temperatures in 19.3.3 Empirical Models—Purely empirical models de-
solutions with constant pH values and chloride ion contents. scribe the observed material responses and dependencies on
Likewise, tests to determine the effects of pH and chloride ion variables without reference to a mechanism. Purely empirical
content would be conducted at various pH values or chloride models appear frequently in the technical literature to quantify
ion contents, and at constant temperature and chloride ion trends in material behavior. These models often serve as a first
content or constant temperature and pH, respectively. Confir- step towards the development of a mechanistic model to
mation of the model could be achieved by comparing the represent the observed trend. An empirical model can have the
measured values and model results under particular conditions same mathematical form as mechanistic and semi-empirical
of temperature, pH, and chloride content that were not used to models; the difference is the functional dependencies on the
determine the functional relationships. Distinctions should be variables (denoted as g) are not based on theory or a mecha-
made between uncertainties that arise regarding the form of the nistic model, Eq 3:
model, the precision and bias in the test data, and the fitting
constants that are extracted from the test data to be used in the Y 5 g ~ x i ... x n ! 1ε (3)
model to properly evaluate the total uncertainty in the model The residual term ɛ represents the difference between the
results (see Section 24). response calculated by using the mathematical expression and
19.3.2 Semi-Empirical Model—Several factors may pre- the measured response Y. The coefficients are determined by
clude development of a purely mechanistic model: (1) The time optimizing the responses in characterization tests conducted
and resources required to develop such a model may be using values of the variables that span their ranges in the
impractical. (2) An analytical representation of the alteration service condition to minimize the residual. The mathematical
behavior may not be possible. (3) The relationships may be so form of an empirical relationship between the measured
complex that numerical solutions using the model might not be responses and variable values may provide insight into mecha-
feasible, even with the fastest computers available. Thus, a
nisms that may control the alteration. For example, the
purely mechanistic model may be unwarranted, impractical, or
observation of a dependence on the square root of test duration
unattainable.
may be indicative of control by a diffusion process with
19.3.2.1 A semi-empirical model uses mechanistically-
constant diffusivity.
based terms for some processes, while other processes are
represented by terms based on empirical observations. Semi- 19.3.3.1 The approach for empirical models is to determine
empirical models represent a practical compromise between a relationship that is consistent with or provides an upper
mechanistic and empirical models. These models are illustrated bound to observed data within an acceptable margin. A model
mathematically by Eq 2: is considered to be purely empirical when a mechanistic
relationship between the variables and response cannot be
Y 5 f ~ x i … x n ! 1ε, (2) postulated or inferred. The correlation between the variable
where Y is the dependent variable measured by a test and the response is analyzed empirically to determine a
response and xi through xn are the independent variables that possible functional relationship. The independent variables that
have been identified to affect Y. The term f(xi… xn) represents affect a particular response may initially be chosen on the basis
a plausible but inexact functional expression (or set of expres- of judgment, inconclusive data, or some partially applicable
sions) for the relationship between the independent variables theories. Other variables may become apparent during testing.
and the measured test response. The functional expressions are For example, it might be hypothesized that the corrosion rate of
usually determined by evaluating the results of attribute, a certain steel should be affected by temperature, pH, chloride
characterization, and accelerated tests that isolate or highlight [Cl-] concentration, and Eh of the water to which it is exposed.
the effect of a particular variable. The term ɛ is a constant A possible conceptual model could have the following math-
residual value included in the expression because the function ematical form, Eq 4:
f(xi… xn) may not fully represent the dependence of the test dY/dt 5 F B~ Eh! A ~ T ! P ~ pH, Cl2 ! 1ε (4)
responses on the set of variables. This may because it is not
possible to determine a functional relationship (either mecha- where dY/dt is the corrosion rate (for example, mil/y), B(Eh)
nistic or empirical) between some variables and the measured is a function relating the measured corrosion current to the
responses, because not all variables are known, because the solution Eh, F represents the constants in Faraday’s Law, A(T)
effects of some variables may not be distinguishable, etc. In is the temperature dependence, and P(pH, Cl-) is a function
many cases, the effects of more than one variable are lumped relating the catalyzing and inhibiting effects of pH and Cl- (for
together and represented by a single model variable. example, on the formation of a passive layer), and ɛ represents
19.3.2.2 The approach for developing a semi-empirical the residual between the measured rate and the model due to
model is to postulate a series of steps or reactions as being approximations and processes not taken into account.

14
 C1174 − 17
19.3.3.2 The functional forms determined in empirical mod- 20. Development of a Materials Behavior Model
els may only be applicable under the test conditions used to
20.1 Model development is iterative with testing. As indi-
generate the data. That is, the values of unidentified variables
that are taken into account by the residual may be different cated in Fig. 1, the initial step is to formulate conceptual
under different test conditions. In the example in 19.3.3.1, the models for the materials alteration modes that were expected to
rate may depend on the chromium content of the steel. The be most important in the problem definition stage. The initial
composition of the steel may be taken into account in the value conceptual model may be a simplification of the overall
of B, P, or ε in the rate expression depending on how the material alteration behavior or may address a particular process
electrochemical and chemical processes are affected. that contributes to the overall mechanism. For example, it may
19.3.3.3 Consider the case where the steel corrosion rate be postulated that components are released from a material into
depends on the solution Eh according to the Butler-Volmer solution by a two-stage process of oxidation and dissolution
model. In that case, the Eh-dependence in the empirical steps. Separate models may be developed and assessed for each
function B(Eh) can be represented using the Butler-Volmer stage. The possible impact of neglecting some alteration modes
equation, which also accounts for the temperature dependence as the conceptual model is developed must be assessed and
of the oxidation reaction. Replacing the B(Eh) term with the considered as potential uncertainty in the model. The concep-
Butler-Volmer equation will affect the A(T) term and probably tual model is used to identify information needs and to plan
also the P and ɛ terms. Additional characterization tests may be tests to acquire the test data required to use or evaluate the
required to determine lumped parameter values for different model. These will include attribute, characterization, service
representations. condition, and accelerated tests. Fig. 3 shows the modeling

FIG. 3 Recommended Procedure for Developing Accelerated Tests for Materials

15
 C1174 − 17
process in more detail. Depending on the level of mechanistic manner (see Section 27). Preliminary tests and analyses used to
understanding of the alteration processes, a model may be develop initial conceptual models do not need to be qualified.
considered empirical, semi-empirical, or mechanistic. 20.3 Data may be rejected on the basis of inadequate test
20.1.1 Empirical analysis of the conceptual model is usually controls or on an objective basis, such as statistical analysis to
the initial step taken because the significant variables are identify outliers. Data that are not fully qualified may be used
generally unknown or uncertain. In this case, the data from if they are the only data available that address a particular
service condition and characterization tests, and possibly from issue, are adequate for their intended use in formulating the
other sources (for example, attribute tests and natural analogs) model, and conclusions drawn from them are assigned an
are analyzed to identify relationships and trends in the data. appropriate degree of uncertainty. Data not originally devel-
The consistency between the expected behavior from the oped under the required QA program may potentially be
conceptual model and the observed trends in the data is used to qualified for use in model and parameter development and
evaluate the adequacy of the model and provide insights to validation as allowed by procedures of the implementing
modify the model as necessary. organization.
20.1.2 Another objective of empirical analysis is to look for
evidence of changes in the relationship between the indepen- 20.4 Confidence in the empirical model can be provided
dent variables and the test response as the values of test using analyses methods such as Expert Elicitation.10
variables (for example, temperature and pH) are changed. This
21. Model Validation (Support for and Confidence in
may indicate a change in the alteration mechanism and is
Models)
particularly important for the analysis of accelerated test
results. Identification of trends in the data during empirical 21.1 Model validation is the process in which model results
analyses may lead to hypotheses of mechanistic relationships. are compared with independent measurements or analyses.
The conceptual model may be modified to take these relation- Validation provides support for and confidence in the applica-
ship into account and other experiments designed to test the tion of the model to acceptably estimate the alteration behavior
hypotheses. The empirical conceptual model may thereby for conditions that cannot be tested directly. In supporting the
evolve into a semi-empirical model.
20.2 All data used to develop the final process models and 10
Kotra, J. L., Lee, M. P., Eisenberg, N. A., and DeWispelare, A. R., “Branch
determine model parameter values important to waste isolation Technical Position on the Use of Expert Elicitation the High-level Radioactive
should be collected in a Quality Assurance (QA)-approved Waste Program, NUREG-1563, US NRC, Washington, D.C., 1996.

FIG. 4 Details of “Perform Modeling” Module in Fig. 1

16
 C1174 − 17
material alteration models developed using the techniques be compared to the known compositions of U-bearing mineral
described above, it should be recognized that “validation” (or phases known to occur in the repository environment to
proof in the traditional sense) in terms of comparison of model support the aspect of the model representing the effects of
result with a material response measured over the full range of alteration phases. However, confidence in the model based on
expected in-service conditions is obviously impossible when analog information will typically have some inherent limita-
one of the key conditions is the post-closure time period of tions that must be acknowledged when documenting the model
thousands of years. Instead, support for and confidence in support. For example, the analog information described here is
model results is based on comparisons of the results of limited because the naturally-occurring uranium phases did not
in-service condition, in-situ, and confirmation tests and analy- evolve in close proximity with other materials that will be
sis of analogs—making allowance for the long time periods present in the EBS, such as zirconium cladding and stainless
and modeling uncertainties—is the general standard that the steel containment materials.
models should be required to meet. Thus, the term ‘validation’
is used sparingly in this practice and when used is referring to 21.4 In cases where there are limited independent data or
those activities that provide support for and confidence in analyses to adequately support a materials behavior model, a
models used for estimating the performance of materials for bounding analysis can be used. A model that can be shown to
geologic disposal applications. bound the rate of alteration under all credible environmental
21.1.1 Support for and confidence in models is enhanced conditions may be regarded as acceptable for the purposes of
when multiple lines of evidence are provided (for example, its usage, which would generally be a conservative over-
laboratory test, in situ tests, analog tests). estimation of the rate of alteration. The bounding model could
be mechanistic, semi-empirical, or empirical with regard to the
21.2 The models are generally derived using data limited to process being bounded.
tests conducted for durations that are very short compared to 21.4.1 An alternative approach would be to perform analy-
the very long times to which the models will be applied. Many ses that show there is an upper bound to the amount of
material behavior properties do not depend on time directly. alteration due to limits imposed by the mode of alteration. If
Instead, they depend on environmental conditions that may this is the case, then a constant value could be used for the
change over time in the disposal system within expected ranges alteration rather than a model that depends on the values of
(for example, temperature, mechanical loads, and pH). Al- environmental variables. For example, the near-field tempera-
though the values of environmental variables will vary within ture in the repository will eventually decrease as a function of
expected ranges over time, the dependence of the material time. If the bounding temperature is chosen to be the maximum
response on those variables will not change unless the mecha- temperature, then the need to model the variability of the
nism changes. Confidence in models for these processes can be process with temperature might be eliminated. This option is
enhanced by conducting tests under conditions that span the applicable only if the bounding values used for the relevant
full range of environmental conditions anticipated to occur parameters can be justified and demonstrated to provide upper
over the long service life of the disposal system. Confidence in bounds. For example, if a reaction product that retards the
the predictive capacity is usually higher for mechanistic alteration process forms at some maximum temperature but
models than for empirical models because of the relationships does not form at a lower temperature and the process is not
between the variables and the test response can be attributed to retarded, then use of the maximum temperature might not yield
specific processes. However, the same test data can often be the bounding degree of alteration and is therefore not a
interpreted using different models that predict different long- justifiable bounding value. A thorough evaluation of the
term behavior. For example, although the dissolution rates of bounding conditions chosen and the effect of these conditions
borosilicate glasses are known to depend on temperature, pH, on the reaction process should be conducted before using the
and the activity of dissolved silica, there is an on-going debate bounding condition.
whether the dissolution rate is controlled by surface dissolution
reactions or diffusion through surface layers. Dissolution rates 21.5 Support for and confidence in material behavior mod-
measured over the full range of temperatures, pH values, and els may also be provided by the use of accelerated tests. For
silica concentrations (up to saturation concentrations) are example, a waste container material could be exposed to water
represented equally well with the two mechanisms. or water vapor at a temperature higher than the anticipated
Furthermore, changes in the mechanism may occur as altera- in-service condition. The corrosion product resulting from the
tion progresses that cannot be predicted by a mechanistic test could then be compared to that estimated by the model for
model, such as the nucleation of a secondary phase that affects in-service conditions, and, if similar, could be used to provide
the glass dissolution rate. In this case, the model providing the support for and confidence in the corrosion model as providing
higher upper bound may be preferred to provide the more an upper bound for the long-term repository conditions.
conservative analyses. 21.6 It should be recognized that all models are essentially
21.3 Support for and confidence in some materials behavior simplified representations of actual alteration processes. Mod-
models may be obtained using natural analogs. For example, an els developed under the foregoing procedures may be super-
alteration model for the degradation of commercial spent fuel seded by models that better represent the process or are more
might be based on test data in which mineral phases formed as efficiently implemented. Should a new model give results that
a result of the dissolution of uranium dioxide generating conflict with the results obtained from the initial model, the
saturated solutions. The composition of these phases can then new model must be supported by comparing model results with

17
 C1174 − 17
the same test data used to validate the previous model. 22.2 Estimates of the performance of the EBS materials
Additional tests may be required to discriminate between used for geologic disposal consider time scales over much
alternative models. longer time periods (orders of magnitude) relative to the length
of time tests are conducted to develop the models and
21.7 If no representation model can be developed that is determine model parameter values. In some cases, (for
consistent with the test data, it may be necessary to return to example, corrosion of stainless steels), repository performance
the Problem Definition stage (see Section 9). The appropriate- calculations will be made using material behavior models
ness of the test data should be re-evaluated. If no alternative based on a range of future environments.
models can be conceptualized, it may be necessary to exit the 22.2.1 If appropriate analogs are available, however, the
process and select another course of action. Such options are models are used to interpolate between existing data in order to
outside the scope of this practice. The Swedish and United estimate the materials behavior. Since precise matches of
States regulatory authorities (SKI and NRC) have provided analog compositions are unlikely, models must also serve to
regulatory perspectives on model validation in the high-level extrapolate or, preferably, interpolate data against material
radioactive waste management programs. composition in these instances. The intent of using analog
ESTIMATING PERFORMANCE materials is to increase the confidence in the performance
estimates; the models used for extrapolation or interpolation
should both adequately represent available data and capture the
22. Scope
extent of mechanistic understanding of alteration processes for
22.1 This element describes the recommended procedure each material. However, further confidence is afforded the
for using models to estimate materials behavior for perfor- performance estimates when they are based on interpolations
mance assessment purposes. Generally, there will be two broad of available data.
categories of models used in estimating performance of mate- 22.3 Post-Closure Performance Time—Repository perfor-
rials behavior. A first category of model are ‘process-type’ mance for the post-closure period requires calculations to
models used to represent individual EBS materials under represent the effects of credible features, events, and processes
in-service conditions in a detailed manner to support the on material alteration over very long times (for example, 10
identification of important processes and parameters affecting 000 years and longer).
their alteration. These models are based on testing during 22.3.1 For some process models, the effect of time on
development and confirmation activities. Although these mod- material alteration behavior will occur through changes in the
els could be used directly in performance assessments, these environmental conditions that are variables in the model, such
process-type models tend to be more detailed than is required as temperature, pH, and solution chemistry, rather than being
in performance assessments that represent the overall reposi- an independent variable.
tory system. A second category of models are those that are 22.3.2 Modelling may also require the evaluation of inter-
further simplified (based on the identified important processes actions between various materials in the repository system.
and parameters) for use in performance assessment models Some of these interactions may be taken into account through
because they are more efficient for use in probabilistic evalu- variables that are included in behavior models of the individual
ations of overall performance. These ‘performance assessment’ materials while other interactions may require additional vari-
models are used to represent the alteration behaviors of the ables or expanded ranges of values. For example, the dissolu-
EBS materials having the greatest impact on system tion of high-level waste glass and concrete EBS components
performance, as identified via the ‘process-type’ models. will likely cause an increase in the pH of groundwater
Model development is an iterative process and it can be contacting the steel waste package components to values
expected that understanding and information learned with higher than expected for local groundwater. If the model for
either the process models or the performance assessment steel degradation has a pH-dependence term, then the range of
models can be used to assist development and support for both pH values for which the steel degradation model is supported
models. However, this SP is applicable more to the process- should include the higher pH values expected to be generated
type model that represents a smaller subset of the overall by glass or cement degradation.
performance model and generally has a limited set of processes 22.3.2.1 Reactions between the breached EBS barrier ma-
that can be evaluated in material testing. terials and the ground water may continue to be important over
22.1.1 The process models can be used to compare mea- time because the resulting modification of the ground water
sured and predicted values at several stages in the logical composition may affect the alteration of spent fuel and glass.
progression shown in Fig. 1. It is useful to differentiate
between the two distinct purposes of these results: material 23. Environmental Conditions
behavior estimates are used to identify significant processes 23.1 It is recognized that environmental conditions to which
and parameters that require additional testing support; perfor- materials will be exposed in the repository may change after
mance estimates for waste forms and EBS materials are used emplacement. Estimates generated from most materials behav-
iteratively to identify changes to the EBS design and materials ior models will depend on the environmental conditions that
selection that would improve performance model predictions are considered, since most models will include dependencies
and repository performance under the applicable and relevant on temperature, groundwater chemistry, etc. For some behavior
environmental conditions of the repository. models, the change in the environmental conditions will be the

18
 C1174 − 17
primary effect of time on material alteration. The performance 24.1.5 The test data used to identify key variables and
of the EBS materials should be evaluated for the relevant parameter dependencies, and
environmental conditions and, as appropriate, alternative envi- 24.1.6 The environmental service conditions.
ronmental conditions for low-probability scenarios should also
be evaluated. 24.2 Uncertainty in the Mathematical Form of the Model—
The mathematical form of the model is a source of uncertainty
23.2 The time dependence of each environmental variable, whose significance depends on the particular model. The
for example, temperature, groundwater composition, humidity,
uncertainty likely decreases as models become more mecha-
mechanical loading, etc. can be used as an input variables for
nistic and less empirical. Mechanistic models that represent
materials behavior models to represent the material perfor-
known physical or chemical processes of materials alteration
mance as the repository conditions change. Which variables
mathematically have less inherent uncertainty than do empiri-
are included in a particular model is determined during
cal models that represent measured responses. Fully empirical
development of each model.
models may be used as bounding cases, or when mechanistic
23.3 Particular attention should be paid to mutually exclu- models are either not available or not practically achievable,
sive repository conditions to avoid unrealistic environmental but the behaviors predicted using empirical models are consid-
conditions. For example, materials alteration would be pre- ered to have greater uncertainty. For example, a bounding
dicted to be rapid in liquid water at a high temperature (for empirical model based on a measured alteration rate could be
example, above 100°C); however, if the repository is porous used with confidence as part of a conservative analysis of EBS
and thus incapable of maintaining pressurization, these two performance, but confidence in the accuracy of the estimated
elements are mutually exclusive. performance would be very low.
24. Uncertainties in Model Estimations 24.3 Test Data Uncertainty—Essentially all data used to
24.1 General Treatment of Uncertainties—There will be develop models will be obtained over short periods of time
inherent uncertainties associated with characterizing and mod- compared with the repository post-closure time period.
eling long-term behavior of materials that provide barriers to Additionally, the accuracy of all test data used to support
radionuclide release under the disposal environment(s). Esti- model development will have limitations that must be factored
mating the confidence in the long-term behaviors predicted for into any model parameter value derived from that data. This
these materials requires identifying the sources of uncertainties will impact both mechanistic and empirical alteration models.
in each process-type alteration model and how they are 24.3.1 Model Parameter Uncertainty—Model parameter
represented in performance assessments. Quantification of values are generally obtained from test data using data regres-
these uncertainties is most important for those process models sion techniques. These are often used as coefficient values in a
that contribute significantly to the model predictions. This materials alteration model. Alternatively, model parameter
could be done, for example, by first sampling from a probabil- values may be based on theory, data from the open literature,
ity distribution of parameter values for each material alteration expert judgment, or some combination thereof; each approach
model used to calculate the overall repository performance. has associated uncertainty. The uncertainty can be reduced to
Uncertainties for each parameter value could then be statisti- the extent that the alteration mechanism is understood and
cally propagated to derive a quantitative estimate of the overall represented by the analytical model. Model fitting parameters
uncertainty in the calculated performance. The use of alteration cannot provide a degree of accuracy to the alteration model that
models developed to provide reliable estimates of material exceeds the accuracy of the test data from which they are
behaviors over long time periods will be strongly dependent on derived. For example, if corrosion rates are derived from data
the uncertainties in those models. Model uncertainty can with an experimental accuracy of 610 %, the uncertainty in a
originate from uncertainties in the conceptual models, their fitting parameter based on those rates, such as the temperature
mathematical representations, and the data used to determine dependence, will be ≥ 610 %
the dependencies on the variables. Uncertainties in the math- 24.3.2 Propagation of Data Uncertainties—The combined
ematical model arise from the simplifying assumptions and uncertainties in the data and parameters on which the materials
approximations used in formulating the mathematical form of behavior models are based should be evaluated using appro-
the model and the environmental dependencies. Uncertainties priate statistical techniques. These may include propagation
in the conceptual model arise from the incomplete understand- and uncertainty budget approaches.
ing of the processes that contribute to the material behavior and
long-term service conditions. The uncertainties that require 24.4 Uncertainties in Establishing Environmental Service
consideration may be associated with the following: Conditions—Uncertainties in the environmental conditions to
24.1.1 Data which the process model is developed to which materials will be exposed—including the evolution of
represent. those conditions with time and materials interactions—should
24.1.2 The mathematical form of the model itself (for be evaluated for their contributions to the uncertainty in the
example, have appropriate mathematical functions been se- modeled performance of the EBS materials. The evolution of
lected to model the processes?), the physical/chemical environment over the very long post-
24.1.3 The alteration modes represented by the model, closure service time is beyond the scope of this practice, but
24.1.4 Materials interaction effects represented or taken into should be expected to contribute additional uncertainty to the
account in the model, calculated performance.

19
 C1174 − 17
24.5 Confidence in Materials Alteration Estimates— provide information to assist the prioritizing the need for
Estimates of materials behavior over short periods are expected investigation and understanding of the performance of EBS
to have high confidence levels, since models are expected to materials.
reproduce the alteration levels that have been directly observed
and on which they are based. Estimates for longer periods of PERFORMANCE CONFIRMATION
time have lower confidence levels. However, confidence in 25. Scope
model results will also depend on the particular repository
conditions under consideration and the selection of EBS 25.1 Performance Confirmation Requirements—During the
materials. pre-closure or operational period for a geologic repository,
which is anticipated to last for decades prior to permanent
24.5.1 The selection of EBS barrier materials could be
closure of the facility, it is expected that additional data
influenced by the level of confidence they provide to the model
concerning the long-term behavior of EBS materials will be
for the expected primary degradation mode/mechanism. For
obtained through confirmation testing. Confirmation tests are
example, greater confidence in the degradation model for a less
intended to provide confidence in the models developed to
corrosion resistant material X compared to the model for more
represent the performance of the EBS materials during the
highly corrosion-resistant candidate materials may justify in-
post-closure period. The confirmation tests will provide addi-
corporating material X as a “corrosion allowance” into the
tional data to which the process model results can be com-
barrier design.
pared. Confirmation testing should focus on the alteration
24.6 Confidence with Respect to Excluded Alteration behaviors of those EBS materials that are most likely to impact
Mechanisms—The high-level nuclear wastes to be disposed in the overall repository performance. These key materials can be
a repository may consist of many different types of waste identified through performance assessment sensitivity
forms: several kinds of commercial light water reactor spent analyses, expert judgment, analysis of natural analog materials,
fuel assemblies; engineered high level radioactive waste etc. The performance assessments should take into account the
glasses, glass-ceramics, ceramics, alloys; Pu immobilized in risk from potential radiation exposure due to releases from the
glass or ceramics; and several hundred distinct forms of repository based on the inventories of spent fuel and high level
non-commercial and test reactor spent fuels. It is not practical waste to be emplaced and their expected waste degradation
to develop alteration models for all waste form types. It is rates.
expected that the alteration mechanisms for several waste
25.2 Performance Confirmation Testing—Performance con-
forms will be similar enough to the alteration model developed
firmation encompasses a continuous, broad-based, technical
for a waste form that has undergone appropriate testing and
program of tests, experiments, and analyses conducted to
thus can be applied to other waste forms by assuming the same
provide the information needed to confirm the design and
degradation processes are operative. It is possible that an
performance of the repository system during the post-closure
alteration mechanism that was not observed in the testing of
period. It is anticipated that confirmation testing would be
one waste form could significantly contribute to the alteration
performed after the initial development of process and perfor-
of a different waste form under long-term repository condi-
mance models and up to the time of permanent closure of the
tions. An emphasis on the mechanistic understanding of
repository.
potential alteration modes, careful selection of representative
25.2.1 A program for monitoring the condition of the waste
materials for testing, and appropriate characterization and
packages may be established at the geologic repository opera-
accelerated tests should minimize the possibility that a signifi-
tions area. Waste packages chosen for the confirmation test
cant alteration mode will be overlooked or unduly discounted
program must be representative of those to be emplaced in the
when developing the alteration models. Using materials repre-
underground facility.
senting the range of waste form for Confirmation testing (see
25.2.2 Consistent with safe operation at the geologic reposi-
Section 25) would add confidence that no reasonably probable
tory operations area, the environment of the waste packages
alteration mode has been overlooked.
selected for the waste package monitoring program (WPMP)
24.7 Uncertainty in Performance Assessment—Probabilistic must be representative of the post-closure environment in
performance assessment sometimes use a bounding or conser- which the wastes are to be emplaced.
vative approach. The bounding approach uses a worst case 25.2.3 The WPMP should include laboratory experiments
value from a range of uncertain values. The conservative that focus on the internal condition of the waste packages. To
approach is not worse case and adopts conservative values the extent practical, the environment experienced by the waste
rather than using realistic values that may be very uncertain. packages emplaced within the underground facility (for
Both approaches can be incorporated in the uncertainty example, underground research laboratories) during the WPMP
distribution, if necessary. These two approaches are used as must be duplicated in the laboratory experiments.
long as the safety objectives of the disposal system continue to 25.2.4 The WPMP should continue as long as practical.
be met. Model representations of the degradation processes are
generally of a simple form in the performance assessment 26. Specific Procedure
computer code. This relies, in part, on bounding and conser- 26.1 Identify processes and parameters that are important to
vative approaches for the simplicity but allows the perfor- post-closure performance. Identification should be made based
mance assessment results to take the significant uncertainties on a risk-informed performance-based (RIPB) approach. RIPB
into account. Thus, the performance assessment can be used to focuses on tests, experiments, and analyses that address

20
 C1174 − 17
Features, Events, and Processes (FEPs) that are significant to ization (ISO), and other standards should be used as guidance
the performance of the EBS materials for the repository or references for data collection and modeling.
conditions. 27.3 Acceptable data must be recoverable, defensible, and
26.2 Select processes and associated parameters that require traceable.
performance confirmation testing using RIPB approach. 27.3.1 Data are recoverable when they are completely
documented in accessible records.
26.3 Analyze existing data and models to establish toler-
27.3.2 Data are defensible when they have been obtained by
ances or limits or deviations from values for key parameters of
documented and approved test methods using good laboratory
the selected processes for the expected performance of the EBS
and field test practices and are reproducible.
materials.
27.3.3 Data are traceable when they can be related through
26.4 Identify completion criteria and guidelines for correc- an unbroken chain to acceptable reference standards, calibra-
tive actions to be applied if variances occur. tion checks, and parallel experiments using standard reference
26.5 Conduct detailed planning of test and monitoring materials from authoritative sources such as standards bodies
activities to measure key parameters. and institutional standards organizations.

26.6 Monitor performance, perform tests, and collect data. 27.4 Models in the form of computer software must be fully
documented as defined by national law and regulation and a
26.7 Analyze and evaluate the collected data using process software quality assurance plan approved under the QAP
models, statistical tests, and total system performance assess- governing the activity.
ments.
28. Precision and Bias
26.8 Recommend and implement appropriate actions if data
are outside the established tolerances or limits or deviate from 28.1 The parameter values in the alteration models devel-
values of the parameters relevant to the expected performance oped under this practice, when determined using curve-fitting
of the EBS materials. and regression of experimental data from accelerated,
characterization, and service condition tests, should reflect the
precision and bias limitations of that data. The accuracy of a
27. Quality Assurance
materials alteration model should not be taken as greater than
27.1 This practice covers activities related to the evaluation the precision of the test data from which the model, model
of the long-term behavior of materials used in the EBS for parameters, and model parameter values are derived. State-
geological disposal that are subject to the quality assurance ments of precision and bias should be developed for the test
requirements defined by national law and regulation. data used to support model development and the consequent
27.2 All data collection and modeling shall be done under a quantitative performance results from the application of this
qualified Quality Assurance Program (QAP) defined by na- practice. (See Practices E177, E178, and E583).
tional law and regulation. 28.2 The factors that contributed to the uncertainty in the
27.2.1 The consensus standards such as ANSI NQA-1, model results should be identified and the significance of their
ASTM standards, the International Organization for Standard- contribution described and, when possible, quantified.

ASTM International takes no position respecting the validity of any patent rights asserted in connection with any item mentioned
in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk
of infringement of such rights, are entirely their own responsibility.

This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the
responsible technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should
make your views known to the ASTM Committee on Standards, at the address shown below.

This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959,
United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above
address or at 610-832-9585 (phone), 610-832-9555 (fax), or service@astm.org (e-mail); or through the ASTM website
(www.astm.org). Permission rights to photocopy the standard may also be secured from the Copyright Clearance Center, 222
Rosewood Drive, Danvers, MA 01923, Tel: (978) 646-2600; http://www.copyright.com/

21