ANL-EBS-MD-000003 REV 00 13 January 2000 1. PURPOSE 1.1 ANALYSES AND MODELS REPORT As described in the License Application Design Selection Report, the recommended waste package design is Engineering Design Alternative II (CRWMS M&O 1999a). This design includes a double-wall waste package (WP) underneath a protective drip shield (DS). The purpose and scope of the process-level model is to account for both general and localized corrosion of the waste package outer barrier (WPOB), which is assumed to be Alloy 22 (UNS N06022-21Cr-13Mo-4Fe-3W-2C-Ni [American Society for Testing and Materials (ASTM) 1997a]). This model will include several sub-models, which will account for dry oxidation (DOX), humid air corrosion (HAC), general corrosion (GC) in the aqueous phase, and localized corrosion (LC) in the aqueous phase. This analyses and models report (AMR) serves as a feed to the waste package degradation code (WAPDEG) analyses. It also serves as a basis for the WP process model report (PMR) and model abstraction for WAPDEG (CRWMS M&O 1999b). Lists of Data Tracking Numbers (DTNs) and their Q-status is included in the Document Input Reference System database and are also included in the Technical Data Management System database and are not in this document. 1.2 BACKGROUND ON ALLOY 22 Alloy 22 (UNS N06022) is now being considered for construction of the outer barrier of the WP. This alloy consists of 20.0-22.5% Cr, 12.5-14.5% Mo, 2.0-6.0% Fe, 2.5-3.5% W, 2.5% (max.) Co, and balance Ni (ASTM 1997a). Other impurity elements include P, Si, S, Mn, Cb, and V (CRWMS M&O 1999e; Treseder et al. 1991). Alloy 22 is less susceptible to LC in environments that contain Cl- than Alloys 825 and 625, materials of choice in earlier designs. The unusual LC resistance of Alloy 22 is apparently due to the additions of Mo and W, both of which are believed to stabilize the passive film at very low pH (Hack 1983). The oxides of these elements are very insoluble at low pH. Consequently, Alloy 22 exhibits relatively high thresholds for localized attack. Very high repassivation potentials have been observed by some (Gruss et al. 1998), while others have found very low corrosion rates in simulated crevice solutions containing 10 wt% FeCl3 (Gdowski 1991; Haynes 1987, 1988). Furthermore, no significant localized attack of Alloy 22 has been seen in crevices exposed to water compositions representative of those expected in the repository. Such tests have been conducted in the Yucca Mountain Project’s (YMP’s) Long Term Corrosion Test Facility (LTCTF) (Estill 1998). Test media used in this facility include simulated acidic concentrated water (SAW), which is about one-thousand times more concentrated than the ground water at Yucca Mountain (J-13 well water) and which has been acidified with H2SO4 (Gdowski 1997c). The measured pH of SAW is approximately 2.7. 1.3 ENVIRONMENT The WP will experience a wide range of conditions during its service life. Initially, the highlevel waste containers will be hot and dry due to the heat generated by radioactive decay. However, the temperature will eventually drop to levels where both HAC and aqueous phase corrosion (APC) will be possible. Crevices will be formed between the WP and supports; ANL-EBS-MD-000003 REV 00 14 January 2000 beneath mineral precipitates, corrosion products, dust, rocks, cement, and biofilms; and between layers of the containers. There has been concern that the crevice environment may be more severe than the near field environment. The hydrolysis of dissolved metal can lead to the accumulation of H+ and a corresponding decrease in pH. Electromigration of Cl- (and other anions) into the crevice must occur to balance cationic charge associated with H+ ions (Gartland 1997; Walton et al. 1996). These exacerbated conditions can set the stage for subsequent attack of the corrosion resistant material by passive corrosion, pitting (initiation and propagation), stress corrosion cracking (SCC), or other mechanisms. 1.4 RELATIONSHIP TO PRINCIPAL FACTORS Degradation of the WP is key to understanding one of the most important principal factors in repository performance. This principal factor is the amount of water transmitted into and the rate of release of radionuclides out of the WP. Once water contacts (touches) the surface of the WP, its fate becomes intertwined with that of the WP. The models and supporting experimental data to account for WP degradation, as well as the evolution of water involved in the various degradation processes, have been sponsored by the YMP. These models and supporting experimental data are reported in two companion PMRs, one for the WP and another for the waste form. This AMR addresses the development of the models to account for the degradation of the outer barrier of the WP, based upon data generated by the YMP, and an integral part of the WP PMR. 1.5 ACTIVITY PLANS Approved activity plans and technical development plans were used in the performance of the work described in this document. Any necessary deviations from these activity plans are documented in the corresponding scientific notebooks (SNs). These procedures are compliant to the Office of Civilian Radioactive Waste Management (OCRWM) quality assurance (QA) requirements. 1.6 SUMMARY OF MODEL The model for the GC and LC of Alloy 22 is summarized in Figure 1. The threshold relative humidity (RH) is first used to determine whether or not DOX will take place. If DOX is determined to occur, the parabolic growth law represented by Equations 11 and 13 is then used to calculate the corrosion rate as a function of temperature. If the threshold RH is exceeded, HAC will occur in the absence of dripping water, and APC will occur in the presence of dripping water. If APC is assumed to occur, the corrosion and critical potentials are used to determine whether the mode of attack is general or localized. The correlation represented by Equation 17 and Table 5 can be used as the basis for estimating these potentials at the 50th percentile. Since the material specifications will be based partly on the measured corrosion and critical potentials, it is assumed that these potentials will be uniformly distributed about the 50th percentile values determined from the correlation. For example, the 0th and 100th percentile values of Ecorr are assumed to be at Ecorr (50th percentile) ± 75 mV. This acceptable margin was determined by splitting the differences shown in Table 6. Acceptability is defined as a condition where no LC occurs. Similarly, the 0th and 100th percentile values of Ecritical are assumed to be at Ecritical (50th percentile) ± 75 mV. Material falling outside of these specified ranges will not be accepted. If ANL-EBS-MD-000003 REV 00 15 January 2000 the comparison of Ecorr to Ecritical indicates GC, the distribution of rates determined from the LTCTF will be used as the basis of the GC rate. If the comparison indicates LC, the distribution of rates presented in Table 22 will be used. This model does not yet account for the effects of aging on corrosion rates. However, such enhancements of the corrosion rate will be accounted for in the future. Other correlations of Ecorr and Ecritical data given here may also be used, if deemed appropriate. Figure 1. Schematic Representation of Corrosion Model for Alloy 22 Outer Barrier critical RH RH = ? Dripping DOX dt dp HAC dt dp GC dt dp LC dt dp critical corr E E = C T ° = 100 ) ( ) ( ) ( 3 2 1 T f i T f E T f E SCW pass critical corr = = = ) ( ) ( ) ( 6 5 4 T f i T f E T f E SSW pass critical corr = = = Dripping RH T , , Effective dt dp yes yes yes yes no no no no ANL-EBS-MD-000003 REV 00 16 January 2000 1.7 UNCERTAINTY AND VARIABILITY The primary uncertainty in the threshold RH for HAC and APC is due to the presence of nitrate. Values of the equilibrium RH as a function of temperature for a saturated solution of NaNO3 are given in Table 9 and Figure 8 of the AMR on WP surface environment (CRWMS M&O 2000a). Despite significant experimental work at Lawrence Livermore National Laboratory (LLNL), there continues to be significant uncertainty in the threshold RH for HAC and APC. In an ideal case, the crevice corrosion temperature can be estimated from the intersection of the lines representing the corrosion and threshold potentials at elevated temperature. To force crevice corrosion to occur in the model, Ecorr and Ecritical can simply be equated over temperature ranges of uncertainty (90-120°C). It is assumed that the crevice corrosion temperature is uniformly distributed over this range of uncertainty. Additional data is needed to fill this void. From experimental measurements presented in Section 6.5.2, the maximum uncertainty in the GC rate is estimated to be approximately 6 to 20 nm y-1 in the case of samples with the generic crevice geometry and 11 to 38 nm in the case of samples with the generic weight-loss geometry. These estimates of error are believed to correspond to about one standard deviation (1 s). From the formal error analysis given in Section 6.5.3, it is concluded that the typical uncertainty observed in weight loss and dimensional measurements prevent determination of GC rates less than 38 nm y-1 (~40 nm y-1). Therefore, any measured corrosion rate greater than 160 nm y-1 (4 s) should be easily distinguishable from measurement error. Any rate less than 160 nm y-1 guarantees that the WP outer barrier (wall thickness of 2 cm) will not fail by GC. It is assumed that no scale formation occurs, so all negative rates are eliminated and the entire distribution is assumed to be due to uncertainty. As shown in Section 6.5.2, the rate at the 50th percentile is approximately 50 nm y-1, the rate at the 90th percentile is approximately 100 nm y-1, and the maximum rate is 731 nm y-1. About 10% of the values fall between 100 and 750 nm y-1. 1.8 MODEL VALIDATION The validation process is discussed in Attachment 1, Item 6, of the OCRWM Procedure, AP-3.10Q. Model validation is accomplished in part by comparing experimental measurements of key model parameters to corroborative data that has been published in the open scientific literature. For example, GC rates, corrosion potentials, threshold potentials, and assumed crevice pH values are compared to those published for Alloy 22 and similar alloys in somewhat similar environments (NaCl solutions, sea water, etc.). Validation of the overall model will require extensive review of calculations performed with the abstracted model based upon this processlevel model. That abstracted model is addressed in a companion AMR. Calculated corrosion rates will be compared to experimental measurements to make sure that those rates are reasonable. Absolute validation of a model intended for the prediction of a service life of 10,000 years may not be possible. These models are based upon the best knowledge and insight into these materials and systems available at the present time. As our state of understanding improves, predictions will inevitably be updated to reflect such advancement. Through the implementation of probabilistic calculations that embody the integrated corrosion models provided here, an attempt is made to compensate for our uncertainty as human beings. ANL-EBS-MD-000003 REV 00 17 January 2000 1.9 RESOLUTION OF COMMENTS IN ISSUE RESOLUTION STATUS REPORT The Issue Resolution Status Report (IRSR) recently issued by the Nuclear Regulatory Commission (NRC) provides guidance for the development of process-level models (NRC 1999). The primary consideration in the key technical issues (KTIs) is the container life and source term (CLST). There must be a high degree of confidence in the adequacy of the engineered barrier system (EBS) design, thereby providing assurance that containers will be adequately long-lived, and radionuclide release from the EBS will be sufficiently controlled. The container design and the packaging of spent nuclear fuel and high-level waste glass are expected to make a significant contribution to the overall repository performance. The IRSR defines the physical boundary of the EBS by the walls of the WP emplacement drifts. The IRSR deems six sub-issues to be important to the resolution of the relevant KTI. The first sub-issue is specifically relevant to this AMR, the effects of corrosion processes on the lifetime of the containers. The following are the acceptance criteria for the first sub-issue: 1. The Department of Energy (DOE) has identified and considered likely modes of corrosion for container materials including dry-air oxidation, humid-air corrosion, and aqueous corrosion processes, such as GC, LC, microbial influenced corrosion, SCC, and hydrogen embrittlement as well as the effect of galvanic coupling. Response: This AMR includes process-level models for dry-air oxidation, humid-air corrosion, and aqueous corrosion processes, such as GC, LC, and microbial influenced corrosion. Galvanic coupling effects have been minimized to the extent possible and will be accounted for in greater detail in future revisions. Both SCC and hydrogen embrittlement are dealt with in companion AMRs. 2. DOE has identified the broad range of environmental conditions within the WP emplacement drifts that may promote the corrosion processes listed previously, taking into account the possibility of irregular wet and dry cycles that may enhance the rate of container degradation. Response: This AMR includes environmental thresholds that can be used to switch between dominant modes of corrosion. For example, as the WP temperature drops and the RH increases, the mode of attack changes from dry-air oxidation to humid-air or aqueous-phase corrosion. A comparison of the corrosion and threshold potentials is used to determine whether or not localized corrosion will occur. 3. DOE has demonstrated that the numerical corrosion models used are adequate representations, taking into consideration associated uncertainties of the expected long-term behaviors and are not likely to underestimate the actual degradation of the containers as a result of corrosion in the repository environment. Response: Uncertainties are accounted for in corrosion rates. As shown in Section 6.5.2, the rate at the 50th percentile is approximately 50 nm y-1, the rate at the 90th percentile is approximately 100 nm y-1, and the maximum rate is 731 nm y-1. About 10% of the values fall between 100 and 750 nm y-1. The effects of thermal aging over extended periods of time ANL-EBS-MD-000003 REV 00 18 January 2000 (10,000 years) is being accounted for in the overall corrosion model for the WPOB. This is discussed in detail in Section 6.7 entitled “The Effect of Aging and Phase Instability on Corrosion.” 4. DOE has considered the compatibility of container materials, the range of material conditions, and the variability in container fabrication processes, including welding, in assessing the performance expected in the containers intended waste isolation. Response: The effects of welding and thermal aging on the corrosion resistance of the WP materials will be accounted for as discussed in Sections 5.9 and 6.7 entitled “The Effect of Aging and Phase Instability on Corrosion.” A fully aged sample of Alloy 22 appears to exhibit a less noble corrosion potential, shifted in the cathodic direction by approximately 63 mV in the case of SAW at 90°C, 109 mV in the case of simulated concentrated water (SCW) at 90°C, and by more than 100 mV in the case of basic saturated water (BSW) at 100°C. It is assumed that Ecorr is corrected to account for fully aged material by subtracting approximately 100 mV from values calculated for the base metal. The shift in Ecritical (threshold potential 1) also appears to be approximately 100 mV in most cases. Thus, the difference Ecritical-Ecorr appears to be virtually unchanged. The effect of thermal aging on the corrosion rate is accounted for in the enhancement factor, Gaged, and is based upon a ratio of the non-equilibrium current densities for base metal and aged material. The value of Gaged for base metal is approximately one (Gaged ~ 1), whereas the value of Gaged for fully aged material is larger (Gaged ~ 2.5). Material with less precipitation than the fully aged material would have an intermediate value of Gaged (1 = Gaged = 2.5). 5. DOE has justified the use of data collected in corrosion tests not specifically designed or performed for the Yucca Mountain repository program for the environmental conditions expected to prevail at the Yucca Mountain site. Response: The threshold RH used to determine whether vapor phase attack is by DOX or HAC is based upon the deliquescence point of salt deposits that could form on the WP surface due to aerosol transport. Measurements of GC rates in the vapor and aqueous phases, electrochemical potentials, and other relevant performance data were in test media that can be directly related to water chemistry expected on the WP surface during the service life of Alloy 22. These water chemistries are based upon evaporative concentrations of the standard J-13 well-water chemistry. Crevice chemistry is being measured in situ, with and without the presence of buffer ions. In the aqueous phase, a range of temperature extending from room temperature to 120°C is being investigated. The high-temperature limit is based upon the boiling point of a near-saturation water chemistry without buffer. The expected boiling point of the aqueous phase on the WP surface is expected to be lower. 6. DOE has conducted a consistent, sufficient, and suitable corrosion testing program at the time of the License Application submittal. In addition, DOE has identified specific plans for further testing to reduce any significant area(s) of uncertainty as part of the performance confirmation program. ANL-EBS-MD-000003 REV 00 19 January 2000 Response: The DOE has established a corrosion test program that addresses all anticipated modes of corrosive attack of the WP. Studies include exposure of over 18,000 samples of candidate WP material in the LTCTF. A large number of pre- and post-exposure measurements of dimension and weight allow establishment of distribution functions for representation of the GC rate. Microscopic examination of samples from the LTCTF and other corrosion tests is done with AFM, scanning electron microscopy, X-ray diffraction, Xray photoelectron spectroscopy, secondary ion mass spectrometry, and other state-of-the-art surface analytical techniques. Potentiodynamic and potentiostatic electrochemical tests are conducted with base metal, thermally aged material, and simulated welds. Thermally aged material is fully characterized with the transmission electron microscope (CRWMS M&O 2000b). 7. DOE has established a defensible program of corrosion monitoring and testing of the engineered subsystems components during the performance confirmation period to assure they are functioning as intended and anticipated. Response: The DOE has established a corrosion test program that addresses all anticipated modes of corrosive attack of the WP. There is a clear linkage between the experimental data being collected and modules in the predictive WAPDEG code that serves as the heart of the Total System Performance Assessment. Data and modules have been developed for each key element of the Engineered Design Alternative II design: the WPOB (Alloy 22), the inner structural support (stainless steel 316NG), and the protective DS (Ti Gr 7). Companion AMRs provide data and modules for the stainless steel 316NG and the Ti Gr 7 alloy. ANL-EBS-MD-000003 REV 00 20 January 2000 INTENTIONALLY LEFT BLANK ANL-EBS-MD-000003 REV 00 21 January 2000 2. QUALITY ASSURANCE 2.1 PROCEDURE FOR ANALYSES AND MODELS (AP-3.10Q) The QA program applies to this analysis. All types of waste packages were classified (per QAP- 2-3 REV 10) as Quality Level-1 in Classification of the MGR Uncanistered Spent Nuclear Fuel Disposal Container System (CRWMS M&O 1999c, p. 7). This analysis applies to all of the waste package designs included in the MGR Classification Analyses. Reference CRWMS M&O (1999c) is cited as an example. The development of this analysis is conducted under activity evaluation Long Term Materials Testing and Modeling (CRWMS M&O 1999d) which was prepared per QAP-2-0 REV 5. The results of that evaluation were that the activity is subject to the Quality Assurance Requirements and Description (DOE 1998) requirements. ANL-EBS-MD-000003 REV 00 22 January 2000 INTENTIONALLY LEFT BLANK ANL-EBS-MD-000003 REV 00 23 January 2000 3. COMPUTER SOFTWARE AND MODEL USAGE 3.1 SOFTWARE APPROVED FOR QUALITY ASSURANCE WORK As per AP-E-20-81, raw data for determining the local environment within WP crevices was obtained with a data acquisition system operating with a macro created with LabView Full Development System for Windows 95/NT/3.1 (Serial Number # G10X71724). LabView is considered “industry standard software” and is, therefore, exempt from the OCRWM procedure entitled Software Configuration Management (AP-SI.1Q, Revision 2, ICN 0). The specific application was written and documented by Mr. Richard Green of LLNL in accordance with the first version of the relevant OCRWM procedure (AP-SI.1Q, Revision 0, ICN 0) and is consistent with the later revision of that procedure (AP-SI.1Q, Revision 2, ICN 0). That document is also included in the list of references (Green 1999). Data acquisition software was checked by measuring known quantities. For example, in the case of analog-to-digital converters, known voltage waveforms were measured. Software used with potentiostats to collect CP data was used to measure the voltage-current characteristics across known resistances. Validation was accomplished by ensuring that the application of a given voltage across a known resistance caused a current flow consistent with Ohm’s law. 3.2 SOFTWARE ROUTINES The electronic notebook discussed in AP-E-20-81 was kept with Microsoft Excel 97. Calculations used to manipulate raw data were performed electronically in spreadsheets created with Microsoft Excel 97. The Microsoft Excel 97 that was used was bundled with Microsoft Office 97 Professional Edition for Windows 95/NT or Workstation 4.0 (Serial Number # 269- 056-174). Excel is considered “industry standard software” and is, therefore, exempt from the OCRWM procedure entitled Software Configuration Management (AP-SI.1Q, Revision 2, ICN 0). All spreadsheets have been assigned DTNs, which are listed in the data inventory sheet. Electronic copies of the data inventory sheet and the supporting data are found on the compact disc (CD) read only memory (ROM) and discussed in Section 10. The correlation equations presented in this document were created within Excel spreadsheets. Those correlation equations were checked by hand calculation, using a Hewlett-Packard 20S scientific calculator. All correlation equations were found to be reasonable predictors of the represented data. Many of the tabulated calculations were also checked by hand calculation. For example, the junction potential corrections given in Tables 7 through 11 were checked in this manner and revised as necessary to reflect changes in assumed water chemistry. The error analyses represented by Tables 16 through 20 were also checked by hand calculation. No other significant computational routines are involved in the process level model described here. ANL-EBS-MD-000003 REV 00 24 January 2000 3.3 INTEGRITY OF TRANSFER OF DATA The integrity of electronic data transfer has been verified as required by OCRWM Procedure YAP-SV.1Q. The comparison method was used to ensure the accuracy of the transferred data. A sampling of ~5% of the data in the source file was visually compared to the corresponding data in the transferred file. The data selected was at the reviewer’s discretion. Reviewers included the originator, as well as the document editor and QA staff. ANL-EBS-MD-000003 REV 00 25 January 2000 4. INPUTS 4.1 DATA AND PARAMETERS 4.1.1 Definition of Parameters a dimension of weight loss sample b dimension of weight loss sample c dimension of weight loss sample b0 coefficient in regression equation b1 coefficient in regression equation b2 coefficient in regression equation f(y) probability density function icorr corrosion current density ipass passive current density k parabolic rate constant in DOX model p wall penetration due to corrosive attack t exposure time during weight loss measurement t time in DOX model ui mobility of the ith ion w measured weight loss woxide formula weight of oxide formed during DOX x independent variable in regression equation x oxide thickness in DOX model xi measured parameter in sensitivity (error) analysis xo initial oxide thickness in DOX model y dependent variable in regression equation y computed value in sensitivity (error) analysis zi valence (charge) of the ith ion Ci( a) molar concentration of the ith ion in alpha phase Ci( ß) molar concentration of the ith ion in the beta phase Doxide diffusivity of reacting species through protective oxide Ecorr corrosion potential Ecritical critical potential – threshold for localized attack Ej junction potential – correction for reference electrode junction F Faraday’s constant Gaged enhancement factor for corrosion rate to account for thermal aging of Alloy 22 GMIC enhancement factor for corrosion rate to account for microbial influenced corrosion Joxide flux of reacting species through protective oxide R universal gas constant R2 regression coefficient RH RH RHcritical threshold RH for HAC T temperature ANL-EBS-MD-000003 REV 00 26 January 2000 s standard deviation µ mean . density of Alloy 22 .oxide density of oxide formed during DOX .oxide stoichiometric coefficient for DOX reaction 4.1.2 Determination of Input Parameters Input for this AMR includes bounding conditions for the local environment on the WP surface, which include temperature, RH, presence of liquid-phase water, liquid-phase electrolyte concentration (chloride, buffer, and pH), and oxidant level. The detailed evolution of the environment on the WP and DS surface is defined by a companion AMR entitled Environment on the Surface of Drip Shield and Waste Package Outer Barrier (CRWMS M&O 2000a). This work has been used to define the threshold RH for HAC and APC, as well as a medium for testing WP materials under what is now believed to be a worst-case scenario. This test medium is presented here as simulated saturated water (SSW) and has a boiling point of approximately 120°C. As discussed in the AMR on WP and DS surface environment (CRWMS M&O 2000a), hygroscopic salts may be deposited by aerosols and dust introduced with the backfill and ventilation air. They will be contained in seepage water that enters the drifts and the episodic water that flows through the drifts. Such hygroscopic salts enable aqueous solutions to exist as thin surface films at relative humidities below 100%. The threshold RH (RHcritical) at which an aqueous solution can exist is defined as the deliquescence point (CRWMS M&O 2000a). This threshold defines the condition necessary for aqueous electrochemical corrosion processes of a metal with salt deposits to occur at a given temperature. The deliquescence point of NaCl is relatively constant with temperature and varies from 72-75%. In contrast, the deliquescence point of NaNO3 has a strong dependence on temperature, ranging from an RH of 75.36% at 20°C to 65% at 90°C. The implied equilibrium RH is 50.1% at 120.6°C, the boiling point of a saturated NaNO3 solution at sea level. The primary uncertainty in the threshold RH for HAC and APC is due to the presence of nitrate. Values of the equilibrium RH as a function of temperature for a saturated solution of NaNO3 are given in the AMR on WP and DS surface environment (CRWMS M&O 2000a). It is expected that any other salts with lower deliquescence points (RHcritical) are precipitated in surrounding rock before they reach the WP surface. This threshold obeys the following polynomial in temperature, which is a fit of the data deliquescence point data for NaNO3: RHcritical = -3.5932 ×10-5 × T(°C)3 + 5.9649 ×10-3 × T(°C)2 - 0.45377 × T(°C) + 81.701(Eq. 1) R2 = 0.9854, where R2 is the coefficient of determination and where R is the coefficient of correlation. This correlation is compared to the data in Figure 2. ANL-EBS-MD-000003 REV 00 27 January 2000 y = -3.5932E-05x3 + 5.9649E-03x2 - 4.5377E-01x + 8.1701E+01 R2 = 9.8540E-01 0 10 20 30 40 50 60 70 80 90 0 20 40 60 80 100 120 140 x = T(C) y = Critical RH (%) DTN: LL991208205924.096 Figure 2. Deliquescence Point for Sodium Nitrate Solutions As discussed in the AMR on WP and DS surface environment (CRWMS M&O 2000a), the evaporative concentration of J-13 well water results in the concentration of Na+, K+, Cl-, and NO3 -. J-13 well water has a typical water chemistry for saturated zone and perched waters at Yucca Mountain and a mean composition that was reported by Harrar et al. (1990). During evaporative concentration HCO3 -, Ca2+, and Mg2+ are removed from solution due to carbonate precipitation. The concentration of HCO3 - reaches a constant level, while the concentrations of F- and SO4 2- initially increase but eventually fall due to precipitation. Ultimately, the F- reaches a low steady state value. The SSW used for testing is an abstract embodiment of this observation. The SSW formulation is based upon the assumption that evaporation of J-13 eventually leads to a sodium-potassium-chloride-nitrate solution. The absence of sulfate and carbonate in this test medium is believed to be conservative, in that carbonate would help buffer pH in any occluded geometry such as a crevice. It is well known that polyprotic acids serve as buffers. Experimental data from the scientific and technical literature, the LTCTF and CP measurements, and crevice corrosion experiments at LLNL are used as a basis for this process-level model. The rationale for the test media in the LTCTF is discussed in Section 6.4.2 and by Gdowski (1997a, 1997b, 1997c). Determination of many of the listed parameters is not found specifically in this section but is discussed in detail in Section 6.0, “Analysis/Model.” Specific input parameters from the LTCTF are GC rates from the various test media. CP measurements provide corrosion and threshold potentials necessary for switching from one corrosion mode to another (GC to LC). The crevice corrosion experiments enable the crevice pH to be reasonably bounded. ANL-EBS-MD-000003 REV 00 28 January 2000 Inputs are handled as per OCRWM procedures AP-3.10Q and AP-3.15Q. Data is submitted to the Technical Data Management System and is listed in the associated Data Input Reference Sheet. 4.2 CRITERIA The following criterion applies to general corrosion and localized corrosion of the WPOB of all WP designs (CRWMS M&O 1999f). The disposal container/WP shall be designed, in conjunction with the Emplacement Drift System and the natural barrier, such that the expected annual dose to the average member of the critical group shall not exceed 25 mrem total effective dose equivalent at any time during the first 10,000 years after permanent closure, as a result of radioactive materials released from the geologic repository (CRWMS M&O 1999f) (Section 1.2.1.3). 4.3 CODES AND STANDARDS 4.3.1 Standard Test Media G. E. Gdowski, Formulation and Make-up of Simulated Dilute Water (SDW), Low Ionic Content Aqueous Solution, Yucca Mountain Project, Lawrence Livermore National Laboratory, Livermore, CA, TIP-CM-06, Revision CN TIP-CM-06-0-2, April 4, 1997, Table 1, p. 3. (Gdowski 1997a) G. E. Gdowski, Formulation and Make-up of Simulated Concentrated Water (SCW), High Ionic Content Aqueous Solution, Yucca Mountain Project, Lawrence Livermore National Laboratory, Livermore, CA, TIP-CM-07, Revision CN TIP-CM-07-0-2, April 4, 1997, Table 1, pp. 3-4. (Gdowski 1997b) G. E. Gdowski, Formulation and Make-up of Simulated Acidic Concentrated Water (SAW), High Ionic Content Aqueous Solution, Yucca Mountain Project, Lawrence Livermore National Laboratory, Livermore, CA, TIP-CM-08, Revision CN TIP-CM-08-0-2, April 4, 1997, Table 1, p. 3. (Gdowski 1997c) 4.3.2 Cyclic Polarization Measurements Standard Reference Test Method for Making Potentiostatic and Potentiodynamic Anodic Polarization Measurements, Designation G 5-94, 1997 Annual Book of ASTM Standards, Section 3, Vol. 3.02, pp. 54-57. (ASTM 1997d) Standard Reference Test Method for Making Potentiostatic and Potentiodynamic Anodic Polarization Measurements, Designation G 5-87, 1989 Annual Book of ASTM Standards, Section 3, Vol. 3.02, pp. 79-85. (ASTM 1989) ANL-EBS-MD-000003 REV 00 29 January 2000 4.3.3 General Corrosion Measurements Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens, Designation G 1-90, 1997 Annual Book of American Society for Testing and Materials (ASTM) Standards, Section 3, Vol. 3.02, pp. 15-21. (ASTM 1997e) Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens, Designation G 1-81, 1987 Annual Book of ASTM Standards, Section 3, Vol. 3.02, pp. 89-94, Subsection 8 - Calculation of Corrosion Rate, Appendix X1 – Densities for a Variety of Metals and Alloys. (ASTM 1987) 4.3.4 Comparative Density of Alloy 22 Standard Specification for Low-Carbon Nickel-Molybdenum-Chromium, Low-Carbon Nickel- Chromium-Molybdenum, Low-Carbon Nickel-Chromium-Molybdenum-Copper, and Low- Carbon Nickel-Chromium-Molybdenum-Tungsten Alloy Plate, Sheet, and Strip, Designation B 575-97, 1997. (ASTM 1997a) ANL-EBS-MD-000003 REV 00 30 January 2000 INTENTIONALLY LEFT BLANK ANL-EBS-MD-000003 REV 00 31 January 2000 5. ASSUMPTIONS 5.1 DRY OXIDATION DOX occurs at any RH below the threshold for HAC: critical RH RH < (Eq. 2) This threshold RH for HAC (RHcritical) is assumed to obey Equation 1, which is based upon the AMR entitled Environment on the Surface of Drip Shield and Waste Package Outer (CRWMS M&O 2000a). This process is assumed to result in the formation of an adherent, protective oxide film of uniform thickness. The rate of DOX will be limited by mass transport through the growing metal oxide film. Consequently, the oxide thickness is assumed to obey a parabolic growth law (film thickness proportional to the square root of time). Reasonable values of the parabolic rate constant are assumed as discussed in Section 6.1. DOX is assumed to occur uniformly over each WAPDEG patch, which is comparable in size to that of a LTCTF sample with generic weight-loss geometry. Welding is assumed to have no significant effect on the DOX threshold and rate. Backfill is also assumed to have no significant effect on the DOX threshold and rate. These assumptions are relevant to the analysis presented in Section 6.1. As pointed out in CRWMS M&O (2000a), the deliquescence point can cover a broad range. For example, the deliquescence point of NaOH is 1.63% at 75°C. The deliquescence point of K2SO4 is 97.59% at 20°C. It is assumed that the uncertainty in RHcritical can be represented by a triangular distribution (Section 6.5.4). The value at the 50th percentile is represented by Equation 1. Values at the 0th and 100th percentiles are assumed to be 1.63 and 97.59%, respectively. The specified bounds represent possible binary combinations of anions and cations in J-13 well water. This range addresses concerns regarding a possible lack of conservatism raised during auditing of this AMR. 5.2 HUMID AIR CORROSION HAC occurs at any RH above the threshold: critical RH RH = (Eq. 3) This threshold RH for HAC (RHcritical) is assumed to obey Equation 1, which is based upon the AMR entitled Environment on the Surface of Drip Shield and Waste Package Outer Barrier (CRWMS M&O 2000a). The measured distributions of general corrosion rates for HAC and APC are indistinguishable. Actual rates were below the level of detection. Therefore, the combined distributions presented here are based upon the combined data for the vapor and aqueous phases and are assumed to represent HAC and APC equally well. It is also assumed that the corrosion rate is constant and does not decay with time. Less conservative corrosion models assume that the rate decays with time. HAC is assumed to occur uniformly over each WAPDEG patch, which is comparable in size to that of a LTCTF sample with generic weight-loss geometry. Welding is assumed to have no significant effect on the HAC threshold and rate. ANL-EBS-MD-000003 REV 00 32 January 2000 Backfill is also assumed to have no significant effect on the HAC threshold and rate. These assumptions are relevant to the analysis presented in Section 6.2. 5.3 AQUEOUS PHASE CORROSION At a given surface temperature, the existence of liquid-phase water on the WP depends upon the presence of a salt and mineral deposit. In the presence of such a deposit, a liquid-phase can be established at a higher temperature and lower RH than otherwise possible. In the model discussed here, two conditions must be met for APC, (1) dripping water and (2) RH above the deliquescence point of the deposit at the temperature of the WP surface. While dripping can occur without this condition being met, it is assumed that both conditions are necessary for APC. Without this level of RH, it is assumed that no aqueous phase could be sustained on the surface. critical RH RH = (Eq. 4) This threshold RH for APC (RHcritical) is assumed to obey Equation 1, which is based upon the AMR entitled Environment on the Surface of Drip Shield and Waste Package Outer Barrier (CRWMS M&O 2000a). For the time being, the composition of the electrolyte formed on the WP surface is assumed to be that of SCW below 100°C and that of SSW above 100°C. It is assumed that the corrosion rate is constant and does not decay with time. Less conservative corrosion models assume that the rate decays with time. General APC is assumed to occur uniformly over each WAPDEG patch, which is comparable in size to that of a LTCTF sample with generic weight-loss geometry. Welding is assumed to have no significant effect on the APC threshold and rate. Backfill is also assumed to have no significant effect on the APC threshold and rate. These assumptions are relevant to the analysis presented in Section 6.3. 5.4 DRIPPING CONDENSATE FROM INNER SURFACE OF THE DRIP SHIELD Once the temperature of the DS drops below the dew point, condensation can occur on the inner surface. This condensate can then form droplets that fall through the intervening vapor space and impinge the underlying WP surface, provided that the droplets are sufficiently large so that they can fall through the temperature gradient towards the WP without complete evaporation. After impingement, instantaneous thermodynamic equilibrium is assumed to exist between the condensate and surface deposit. The assumption of instantaneous equilibrium is based on a conservative approach. While much additional work is needed to determine the actual electrolyte composition in this scenario, SCW is assumed below 100°C, while SSW is assumed above 100°C. This assumption is based on the data from CRWMS M&O (2000a), which shows an increase in boiling point as the concentration increases from SCW to SSW. It is assumed that the corrosion rate is constant and does not decay with time. Less conservative corrosion models assume that the rate decays with time. This assumption is relevant to the analysis presented in Section 6.3. 5.5 FLOW THROUGH OPENINGS BETWEEN DRIP SHIELD Section 5.5 is included in this report to show the relationship between water penetrating the protective DS and the water actually contacting the WPOB. Potential ground movement due to ANL-EBS-MD-000003 REV 00 33 January 2000 seismic activity may cause displacement of adjacent DSs along the drift axis, thereby opening pathways that enable dripping water to reach the WP. For a given mass flow of water contacting the outer surface of the DS, the fraction passing through an opening to the WP is assumed to be proportional to the following multiplication factor ( shield T ): shield opening shield A A = T (Eq. 5) where Aopening is the projected area of the opening on the floor of the drift and Ashield is the projected area of the DS on the floor of the drift. If the DS fails due to SCC, a multiplication factor of one is assumed. This assumption is relevant to the analysis presented in Section 6.3. 5.6 THRESHOLD FOR LOCALIZED CORROSION If the open circuit corrosion potential (Ecorr) is less than the threshold potential for localized corrosion (Ecritical), no localized corrosion occurs: critical corr E E < (Eq. 6) Threshold values have been determined for various representative environments, as discussed in Section 6.4. This assumption is relevant to the analysis presented in Section 6.4. As an example, in an ideal case the crevice corrosion temperature can be estimated from the intersection of the lines representing the corrosion and threshold potentials at an elevated temperature. To force crevice corrosion to occur in the model, Ecorr and Ecritical can simply be equated over temperature ranges of uncertainty (90-120°C). It is assumed that the crevice corrosion temperature is uniformly distributed over this range of uncertainty. This assumption is relevant to the analysis presented in Section 6.4.3 and is based on a conservative approach for crevice corrosion temperature. 5.7 EFFECT OF GAMMA RADIOLYSIS ON CORROSION POTENTIAL Effects of oxidant can be accounted for through the open circuit corrosion potential (Ecorr). Based upon published data described in Section 6.5.2, as well as new experimental data shown in this AMR, it is believed that the shift in corrosion potential due to gamma radiolysis will be much less than 200 mV. It is believed that this shift is insufficient to cause LC. This assumption is relevant to the analyses presented in Section 6.4 and 6.5.2. 5.8 EFFECT OF MICROBIAL GROWTH ON CORROSION POTENTIAL The effect of microbial growth on Ecorr and GC rates for Alloy 22 have been studied by Lian et al. (1999) and Horn et al. (1998). End-point measurements of Ecorr in both inoculated and sterile media indicate that microbial growth does not have any large impact on this parameter. The GC rate appears to be doubled in the presence of microbes. More work is needed to help resolve this issue in the future. This assumption is relevant to the analyses presented in Sections 6.4, 6.5, and 6.8 and will be further developed in the future. ANL-EBS-MD-000003 REV 00 34 January 2000 5.9 EFFECT OF AGING AND PHASE INSTABILITY ON CORROSION The WP surface temperature will always be below 350°C, a limit determined by the spent nuclear fuel cladding. By further constraining the WP surface temperature, making sure that it is always below 300°C, the effects of aging and phase instability on the corrosion performance of Alloy 22 can be assumed to be insignificant. An extrapolation of the curves given in the companion AMR on aging and phase stability does not indicate that the phase stability of Alloy 22 base metal will be a problem at less than about 300°C (CRWMS M&O 2000b). However, it must be emphasized that such estimates are preliminary and uncertain. Much additional work is needed in this area. Rebak et al. have investigated the effects of high-temperature aging on the corrosion resistance of Alloy 22 in concentrated hydrochloric acid. However, due to the temperature used to age the samples (922-1033 K) and the extreme test media used (boiling 2.5% HCl and 1 M HCl at 339 K), these data are not considered relevant to performance assessment for the repository. This data will soon be published by R. B. Rebak, N. E. Koon, and P. Crook in an article entitled “Effect of High Temperature Aging on the Electrochemical Behavior of C-22 Alloy.” This paper will appear in the Proceedings of the 50th Meeting of the International Society of Electrochemistry, which documents a conference held in Pavia, Italy, in September 1999. This assumption is relevant to Section 6.4 and 6.5 and will be further developed in the future. 5.10 FLOW THROUGH COINCIDENT PENETRATION IN WASTE PACKAGE It is assumed that the entire mass flow of water passing through the opening in the DS is distributed uniformly on the underlying WP. The fraction of this water that enters a failed WP is assumed to be proportional to the following multiplication factor ( package T ): package failed package A A = T (Eq. 7) where Afailed is the projected area of all failed (completely corroded) WAPDEG patches on the floor of the drift and Apackage is projected area of the WP on the floor of the drift. This assumption is used throughout the analysis. This assumption is based on a conservative approach to allow for maximum water flow. 5.11 ASSUMPTIONS PERTAINING TO INNER BARRIER Section 5.11 is included in this report to emphasize that all corrosion performance is allocated to the WPOB even though the WP is a double-wall container. It is assumed that the stainless steel 316NG inner barrier of the WP provides structural integrity for the WP until the outer barrier fails. No credit is claimed for the corrosion resistance of this stainless steel layer. This assumption is used throughout the analysis and is based on a conservative approach even though the inner barrier is expected to be a barrier for water ingress and radionuclide release. After penetration of the WPOB, the formation of a crevice between the Alloy 22 and 316NG is possible. The formation of a low-pH crevice environment in this interfacial region is possible as ANL-EBS-MD-000003 REV 00 35 January 2000 discussed in Section 6.6.5. Crevice corrosion of the Alloy 22 due to the local chemistry established through hydrolysis of dissolved metal from the 316NG could be severe. While such inside-out attack is not accounted for in the present model, it may be desirable to account for it in the future, especially in crevice regions with welds that might be susceptible to SCC. This assumption is used throughout the analysis. 5.12 QUALIFICATION STATUS OF ASSUMPTIONS The validity of assumptions, and hence the qualification, will be determined through future confirmatory tests. This document and its conclusions may be affected by technical product input information that requires confirmation. Any changes to the document or its conclusions that may occur as a result of completing the confirmation activities will be reflected in subsequent revisions. The status of the input information quality may be confirmed by review of the Document Input Reference System database. ANL-EBS-MD-000003 REV 00 36 January 2000 INTENTIONALLY LEFT BLANK ANL-EBS-MD-000003 REV 00 37 January 2000 6. ANALYSIS 6.1 DRY OXIDATION DOX of Alloy 22 is assumed to occur at any RH < RHcritical, thereby forming an adherent, protective oxide film of uniform thickness. It is assumed that the protective oxide film is primarily Cr2O3. The oxidation reaction is given as (Welsch et al. 1996): 3 2 2 3 2 3 4 O Cr O Cr .. . + (Eq. 8) The rate of DOX is assumed to be limited by mass transport through this growing metal oxide film. Fick’s first law is applied, assuming a linear concentration gradient across the oxide film of thickness x: x C D x C D J oxde oxide oxide . - ˜ . . - = (Eq. 9) where Joxide is the molar flux of the reacting species in the oxide, Doxide is the diffusivity of the reacting species in the oxide, .C is the corresponding differential molar concentration. Oxide growth is related to the flux by: oxide oxide oxide oxide J w dt dx . . × × = (Eq. 10) where .oxide is the stoichiometric coefficient (moles of oxide per mole of diffusing species), woxide is the formula weight of the oxide, and .oxide is the density of the oxide. Integration shows that the oxide thickness should obey the following parabolic growth law (Wagner’s Law [Welsch et al. 1996]), where the film thickness is proportional to the square root of time. This is represented by Equation 11. x = x0 2 + k × t (Eq. 11) where x 0 is the initial oxide thickness, x is the oxide thickness at time t, and k is a temperaturedependent parabolic rate constant. More specifically, k is defined as follows: oxide oxide oxide oxide C D w k . . . × × × × = 2 (Eq. 12) To facilitate an approximate calculation, published values of k can be used (Welsch et al. 1996). From Figure 18 of this reference, it is concluded that all observed values of k fall below a line defined by: log k m2s-1 ( ) [ ]= -12.5 103 T K ( ) . . . . . . - 3.5 (Eq. 13) ANL-EBS-MD-000003 REV 00 38 January 2000 where T is defined as the absolute temperature. The highest temperature is expected to be approximately 350°C (623 K), which corresponds to the limit for the fuel cladding. The value of k corresponding to this upper limit is 2.73×10-24 m2 s-1 (8.61×10-5 square µm per year). After one year, this corresponds to a growth of 0.0093 µm (about 9.3 nm y-1). As will be seen in a subsequent discussion (Section 6.5.3), this estimated rate is comparable to that expected for APC at lower temperatures. It is, therefore, assumed that DOX of the Alloy 22 can be accounted for through application of the parabolic law. The above expression represents a conservative upper bound, based upon the published literature. As discussed in the AMR for corrosion of the titanium DS (CRWMS M&O 2000c), logarithmic growth laws may be more appropriate at relatively lower temperature than parabolic laws. However, such logarithmic expressions predict that the oxide thickness (penetration) asymptotically approaches a small maximum level. In contrast, the parabolic law predicts continuous growth of the oxide, which is much more conservative. Since such conservative estimates of the rate of DOX do not appear to be life limiting and since reliable data for determining the maximum oxide thickness for Alloy 22 do not appear to be available, the parabolic growth law will be used for the WPOB. The DOX model presented here assumes uniform oxidation of the WPOB surface. In the future, the possibility of preferential DOX along grain boundaries in the Alloy 22 should be considered. Such preferential attack would ultimately be diffusion controlled, with the diffusion path being equivalent to the length of oxidized grain boundary. 6.2 HUMID AIR CORROSION HAC is assumed to occur above a threshold RH, provided that there are no impinging drips. critical RH RH = (Eq. 14) This threshold RH for HAC (RHcritical) is assumed to obey Equation 1. The existence of this threshold is due to the dependence of water adsorption on RH. Despite significant experimental work at LLNL, there continues to be significant uncertainty in the threshold RH for HAC and APC. Furthermore, data published by Leygraf (1995) indicates that it may be reasonable to consider HAC at a RH below that predicted with Equation 1 at 20°C. The approximate number of water monolayers on typical metal surfaces as a function of RH is given by Leygraf (1995) and repeated in Table 1. Table 1. Coverage of Metal Surfaces by Water Relative Humidity (%) Number of Water Monolayers 20 1 40 1.5-2 60 2-5 80 5-10 Based upon this data, it might be reasonable to consider the possibility of HAC at only 40% RH. This is the point at which it may be possible for two monolayers of water to exist on the WP ANL-EBS-MD-000003 REV 00 39 January 2000 surface. However, under these conditions there are no electrolytes to facilitate the electrochemical corrosion. As pointed out in CRWMS M&O (2000a), observed deliquescence points cover a very broad range of RH. The deliquescence point of NaOH is 1.63% RH at 75°C, while that of K2SO4 is 97.59% RH at 20°C. It is assumed that the uncertainty in RHcritical can be represented by a triangular distribution. The triangular distribution is described in Section 6.5.4. The value at the 50th percentile is represented by Equation 1. Values at the 0th and 100th percentiles are assumed to be 1.63 and 97.59%, respectively. The specified bounds represent possible binary combinations of anions and cations in J-13 well water. It is assumed that HAC can be treated as uniform GC. The measured distributions of general corrosion rates for HAC and APC are indistinguishable. Actual rates were below the level of detection. Therefore, the combined distributions presented here are based upon the combined data for the vapor and aqueous phases and are assumed to represent HAC and APC equally well. It is also assumed that the corrosion rate is constant and does not decay with time. 6.3 AQUEOUS PHASE CORROSION At a given surface temperature, the existence of liquid-phase water on the WP depends upon the presence of a salt deposit. In the presence of such a deposit, a thin-film liquid phase can be established at a higher temperature and lower RH than otherwise possible. In the model discussed here, it is assumed that two conditions must be met for APC—RH above the deliquescence point of the deposit at the temperature of the WP surface and impinging drips: critical RH RH = (Eq. 15) This threshold RH for APC (RHcritical) is assumed to obey Equation 1, which is based upon the AMR entitled Environment on the Surface of Drip Shield and Waste Package Outer Barrier (CRWMS M&O 2000a). Drips may be due to liquid-phase ground water that flows through openings in the DS or condensate on the underside of the DS. For the time being, the composition of the electrolyte formed on the WP surface is assumed to be that of SCW below 100°C and that of SSW above 100°C. It is assumed that the corrosion rate is constant and does not decay with time. Less conservative corrosion models assume that the rate decays with time. 6.4 LOCALIZED CORROSION 6.4.1 Threshold Potential of Alloy 22 The localized corrosion model for Alloy 22 assumes that localized attack occurs if the open circuit corrosion potential (Ecorr) exceeds the threshold potential for breakdown of the passive film (Ecritical): critical corr E E = (Eq. 16) The repassivation potential is the level at which a failed passive film repassivates, or heals, thereby protecting the surface. Compared to materials proposed for use in earlier WP designs, ANL-EBS-MD-000003 REV 00 40 January 2000 Alloy 22 has superior resistance to localized corrosion. Gruss et al. (1998) have shown that the repassivation potential of Alloy C-22 is far greater than that of Alloy 625, which substantiates this claim (Table 2): Table 2. Repassivation Potentials of Alloys 625 and C-22 Specimen No. Chloride (M) Temperature (°C) Repassivation Potential (V vs. SCE) 625-1 4 95 -0.183 625-2 4 60 -0.167 625-3 1 95 -0.367 625-4 1 95 -0.166 625-5 1 95 -0.153 625-6 1 60 1.001 625-7 0.028 60 0.857 625-8 0.028 60 0.873 C22-1 4 95 0.916 C22-2 4 95 0.911 C22-3 4 95 0.900 C22-4 4 60 0.911 C22-5 1 95 0.829 C22-6 1 60 0.986 C22-7 0.028 95 0.854 NOTE: Gruss et al. (1998) 6.4.2 Cyclic Polarization in Synthetic Concentrated J-13 Well Waters The YMP has used CP to determine threshold potentials for Alloy 22 in test media relevant to the environment expected in the repository. Relevant test environments are assumed to include simulated dilute water (SDW), SCW, and SAW at 30, 60, and 90°C as well as SSW at 100 and 120°C. The compositions of all of the environments are given in Table 3. The compositions of these test media are based upon the work of Gdowski (1997a, 1997b, 1997c). The SSW composition has been recently developed and is being documented in a revision of a companion AMR on the subject of WP and DS surface environment (CRWMS M&O 2000a). The revision is in preparation. In general, anions such as chloride promote LC, whereas other anions such as nitrate tend to act as corrosion inhibitors. Thus, there is a very complex synergism of corrosion effects in the test media. CP measurements have been based on a procedure similar to ASTM G 5-87 (ASTM 1989). Necessary deviations have been noted in the corresponding controlled SNs. Copies of these SNs are maintained by the Management and Operating Contractor (M&O) in Las Vegas. For example, ASTM G 5-87 calls for an electrolyte of 1N H2SO4, whereas SDW, SCW, SAW, and SSW are used here. Furthermore, aerated solutions were used here, unlike the procedure that calls for de-aerated solutions. Representative CP curves are shown in Figures 3 through 9. The shape of these CP curves is categorized as type 1, 2, or 3. ANL-EBS-MD-000003 REV 00 41 January 2000 Table 3. Composition of Standard Test Media Based upon J-13 Well Water Ion SDW SCW SAW SSW (mg/L-1) (mg/L-1) (mg/L-1) (mg/L-1) K+1 3.400E+01 3.400E+03 3.400E+03 1.416E+05 Na+1 4.090E+02 4.090E+04 4.090E+04 4.870E+04 Mg+2 1.000E+00 1.000E+00 1.000E+03 0.000E+00 Ca+2 5.000E-01 1.000E+00 1.000E+03 0.000E+00 F-1 1.400E+01 1.400E+03 0.000E+00 0.000E+00 Cl-1 6.700E+01 6.700E+03 6.700E+03 1.284E+05 NO3 -1 6.400E+01 6.400E+03 6.400E+03 1.310E+06 SO4 -2 1.670E+02 1.670E+04 1.670E+04 0.000E+00 HCO3 -1 9.470E+02 7.000E+04 0.000E+00 0.000E+00 Si 27 (60°C), 49 (90°C) 27 (60°C), 49 (90°C) 27 (60°C), 49 (90°C) 0.000E+00 pH 8.100E+00 8.100E+00 2.700E+00 7.000E+00 NOTE: CRWMS M&O (2000a) A generic type 1 curve exhibits complete passivity (no passive film breakdown) between the corrosion potential and the point defined as threshold potential 1. This interpretation was verified by visual inspection of samples after potential scans and photographic documentation of some of those samples (all samples are held in the archives at LLNL). Threshold potential 1 is in the range where the onset of oxygen evolution is expected and is defined by a large excursion in anodic current. This particular definition of threshold potential 1 is specific to type 1 curves. Type 1 behavior has only been observed with Alloy 22 and is illustrated by Figures 3 and 4. The interpretation of type 1 curves as exhibiting no passive film breakdown is consistent with the ASTM G 61-86. A generic type 2 curve exhibits a well-defined oxidation peak at the point defined as threshold potential 1. Threshold potential 2 is in the range where the onset of oxygen evolution is expected and is defined by a large increase in anodic current. These particular definitions of the threshold potentials are specific to type 2. Repassivation potentials 1 and 2 are defined as the points where the hysteresis loop passes through current levels of 4.27x10-6 and 10-5 amps, respectively (not shown). Repassivation potential 3 is determined from the first intersection of the hysteresis loop (reverse scan) with the forward scan. Type 2 is observed with both Alloy 22 and 316L and is illustrated by Figures 5 through 7. Definitions of the threshold and repassivation potentials are somewhat subjective and may vary from investigator to investigator. Scully et al. (1999) define the threshold potential for crevice corrosion of Alloy 22 as the point during the scan of electrochemical potential in the forward direction where the current density increases to a level of 10-6 to 10-5 A cm-2. Gruss et al. (1998) define the repassivation potential as the point where the current density drops to 10-6 to 10-7 A cm-2, which is comparable to the definition of repassivation potential 3. A generic type 3 curve exhibits a complete breakdown of the passive film and active pitting at potentials relatively close to the Corrosion Potential (Ecorr). In this case, threshold potential 1 corresponds to the critical pitting potential. Type 3 behavior has only been observed with 316L and is illustrated by Figure 8. ANL-EBS-MD-000003 REV 00 42 January 2000 A representative curve for platinum in SCW at 90°C is shown in Figure 9. CP measurements of Pt were made to serve as a basis of comparison for similar measurements with Alloy 22 and other materials of interest. From such comparisons, it is concluded that the anodic oxidation peak observed in type 2 curves (between 200 and 600 mV) is due to an anodic reaction of the Alloy 22 passive film. No anodic oxidation peak is observed in the measurement of Pt. SSW is a saturated sodium-potassium-chloride-nitrate electrolyte, formulated to represent the type of concentrated electrolyte that might evolve on a hot WP surface. This formulation has a boiling point of approximately 120°C at ambient pressure. It is evident in Figure 3 that Alloy 22 maintains passivity at potentials up to the reversal potential (1200 mV versus Ag/AgCl), even under these relatively hostile conditions. In regard to type 2 polarization curves for Alloy 22 in SCW, the electrochemical process leading to the anodic oxidation peak (leading edge at approximately 200 mV versus Ag/AgCl) cannot be determined from the CP data alone. This peak is probably due to some change in the oxidation state of the passive film and probably has very little to do with any loss of passivity. To augment these potentiodynamic measurements, potentiostatic polarization tests have been performed. Figure 10 shows the observed transient current when an Alloy 22 sample is polarized at 200 mV versus Ag/AgCl in SCW at 90°C, close to the potential where the leading edge of the anodic oxidation peak is located. The current initially increases to a maximum of approximately 25 microamps per square centimeter (the sample size is approximately 0.96 cm2) at 9 hours. This corresponds to a typical non-equilibrium passive current density measured for Alloy 22 at this potential in the absence of the anodic oxidation peak. For example, see a type 1 polarization curve for Alloy 22 in SAW. Therefore, in regard to type 2 polarization curves, the anodic oxidation peak does not define any localized corrosion or loss in passivity. Furthermore, threshold potential 1 (leading edge of the anodic oxidation peak at approximately 200 mV versus Ag/AgCl) should not be used as the basis for switching on localized corrosion of Alloy 22. Here, it is also assumed that threshold potential 2 represents the lower bound for breakdown of the passive film. A composite of the CP data is shown in Figure 11. The initial portions of these curves show that passivity is maintained at potentials at least as high as 400 mV versus Ag/AgCl in all cases. The lowest potential at which any electrochemical reactivity of the passive film is observed at approximately 200 mV versus Ag/AgCl. Based upon data presented here, it is concluded that a pitting attack of Alloy 22 should not occur under conditions expected in the repository. To further substantiate this conclusion, it is noted that no pitting of Alloy 22 has yet been observed in samples removed from LTCTF. These data include one-year exposures to SDW, SCW, and SAW at 60 and 90°C. DTNs are associated with Figures 24 through 26. The CP data given in this AMR are for test media believed to be representative of the expected repository environment. In such test media and at plausible electrochemical potentials, it does not appear that there will be significant localized breakdown of the passive film. Furthermore, relatively wide crevices (110 to 540 microns) formed from passive Alloy 22 do not appear to undergo significant increases in hydrogen ion concentration (pH suppression) at reasonable electrochemical potentials. These potentials are generally below the thresholds determined by CP. Finally, Alloy 22 crevices exposed in the LTCTF do not indicate significant crevice corrosion. ANL-EBS-MD-000003 REV 00 43 January 2000 However, it should be noted that the University of Virginia has very recently generated some CP data with very tight crevices and concentrated electrolytes consisting of 5 M LiCl, 0.24 to 0.024 M NaNO3, 0.026 to 0.26 M Na2SO4, and HCl (Scully et al. 1999). Testing was conducted at two temperature levels, 80 and 95°C. The crevices were formed with a multiple crevice former, PTFE tape, and an applied torque of 70 inch pounds. Under these circumstances, some electrochemical activity indicative of crevice corrosion was observed at potentials ranging from 71 to 397 mV versus Ag/AgCl, depending upon the composition of the electrolyte. Using a current density criterion for repassivation of 10-5 A cm-2, repassivation potentials were determined to be slightly above, but relatively close to, the open-circuit corrosion potential. While these concentrated lithium-chloride based electrolytes are not believed to be directly relevant to those conditions anticipated in the repository, the University of Virginia data point out that no attitude of complacency should be adopted in regard to conducting further research in the area of localized corrosion of Alloy 22. Unlike compositions based upon J-13 well water, these electrolytes have no buffer ions per se. Clearly, additional work is needed to better understand the passivity and resistance to localized attack of all WP materials. In the future, similar measurements with test media believed to be relevant to the repository should be conducted. Specifically, testing with the tight-crevice geometry used by the University of Virginia and standard electrolytes such as SDW, SCW, SAW, and SSW should be conducted. As more data become available, the correlations for the corrosion and threshold potentials should be updated, expressing these quantities in terms of temperature, pH, and the concentrations of various ions. The effect of welding and aging should also be accounted for. This AMR should be viewed as works in progress, with each new version reflecting an evolving level of understanding. ANL-EBS-MD-000003 REV 00 44 January 2000 -600 -400 -200 0 200 400 600 800 1,000 1,200 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00 Current (A) Potential (mV vs. Ag/AgCl) Corrosion Potential Passive Non-Equibrium Current Threshold Potential 1 Negative hysteresis loop during reverse scan - no localized breakdown of passive film at reversal potential - no repassivation potential observed Reversal Potential DTN: LL990610105924.074 Figure 3. Type 1 – Alloy 22 in SSW at 120°C (DEA033) -400 -200 0 200 400 600 800 1,000 1,200 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00 Current (A) Potential (mV vs. Ag/AgCl) Corrosion Potential Maximum Passive Current Negative hysteresis loop during reverse scan - no localized breakdown of passive film at reversal potential - no repassivation potential observed Lower Bound for Corrosion Current Threshold Potential 1 DTN: LL990610105924.074 Figure 4. Type 1 – Alloy 22 in SAW at 90°C (DEA002) ANL-EBS-MD-000003 REV 00 45 January 2000 -600 -400 -200 0 200 400 600 800 1,000 1,200 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00 Current (A) Potential (mV vs. Ag/AgCl) Corrosion Potential Lower Bound for Corrosion Current Threshold Potential 1 Positive hysteresis loop during reverse scan - localized breakdown of passive film at reversal potential - repassivation during reverse scan Threshold Potential 2 Repassivation Potential 3 Anodic Oxidation Peak DTN: LL990610105924.074 Figure 5. Type 2 – Alloy 22 in SCW at 90°C (DEA016) -600 -400 -200 0 200 400 600 800 1,000 1,200 1.0E-10 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00 Current (A) Potential (mV vs. Ag/AgCl) Corrosion Potential Lower Bound for Corrosion Current Threshold Potential 1 Threshold Potential 2 Repassivation Potential 3 Positive hysteresis loop during reverse scan - localized breakdown of passive film at reversal potential - repassivation during reverse scan Anodic Oxidation Peak DTN: LL990610105924.074 Figure 6. Type 2 – Alloy 22 in SCW at 60°C (DEA017) ANL-EBS-MD-000003 REV 00 46 January 2000 -500 -400 -300 -200 -100 0 100 200 300 400 500 600 700 800 900 1,000 1,100 1,200 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00 Current (A) Potential (mV vs. Ag/AgCl) Positive hysteresis loop during reverse scan - localized breakdown of passive film at reversal potential - repassivation during reverse scan Threshold Potential 1 Anodic Oxidation Peak Repassivation Potential 3 Corrosion Potential Threshold Potential 2 Lower Bound for Corrosion Current DTN: LL990610105924.074 Figure 7. Type 2 – 316L in SCW at 90°C (PEA002) -600 -400 -200 0 200 400 600 800 1,000 1,200 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00 Current (A) Potential (mV vs. Ag/AgCl) Corrosion Potential Breakdown of the passive film with active pitting Threshold Potential 1 Critical Pitting Potential Repassivation after pit initiation Lower Bound for Corrosion Current DTN: LL990610105924.074 Figure 8. Type 3 – 316L in SSW at 100°C (PEA016) ANL-EBS-MD-000003 REV 00 47 January 2000 0 100 200 300 400 500 600 700 800 900 1,000 1,100 1,200 1.00E-09 1.00E-08 1.00E-07 1.00E-06 1.00E-05 1.00E-04 1.00E-03 1.00E-02 Current (A) Potential (mV vs. Ag/AgCl) Open Circuit Potential DTN: LL990610105924.074 Figure 9. Baseline – Pt in SCW at 90°C (PT001) 0 5 10 15 20 25 30 0 10 20 30 40 50 60 70 80 Time (103 seconds) Current (microamps) Figure 10. Potentiostatic Polarization of Alloy 22 in SCW at 90°C and 200mV Versus Ag/AgCl ANL-EBS-MD-000003 REV 00 48 January 2000 1.0E-13 1.0E-12 1.0E-11 1.0E-10 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 -300 -200 -100 0 100 200 300 400 500 Potential (mV vs. Ag/AgCl) Current (A) ~ Current Density (A/cm2) Current SDW 30 DEA025 Current SDW 60 DEA026 Current SDW 90 DEA027 Current SCW 30 DEA009 Current SCW 60 DEA011 Current SCW 90 DEA010s Current SAW 30 DEA007 Current SAW 60 DEA004 Current SAW 90 DEA002 Current SSW 100 DEA032s Current SSW 120 DEA033s Apparent corrosion current density corresponding to average rates observed in the Long Term Corrosion Test Facility Apparent corrosion rate of 2 microns per year DTN: LL990610205924.075 Figure 11. Alloy 22 in Various Repository Media – Comparison of CP Data 6.4.3 Correlation of Potential Versus Temperature Data for Various Test Media Values of corrosion and threshold potentials are shown in Table 4 and have been correlated as a function of temperature for the conditions of interest. These correlated data are shown in Figures 12 through 15. In general, it has been found that these potential verses temperature data can be represented by the following simple regression equation: x b b y 1 0 + = (Eq. 17) where y is either the corrosion or threshold potential (mV versus Ag/AgCl), and x is the temperature (°C). The linear curves were derived from regression analysis. All correlations are summarized in Table 5, with the correlation for Ecorr and the most conservative correlation for Ecritical labeled. While calculated values of y are believed to have only three significant figures, coefficients in those regression equations used to calculate values of y are given with more figures. By carrying the extra figures during the calculation, round-off error in the final values can be minimized. In the case of type 2 CP curves, the selected threshold potential 1 is determined by the position of the observed anodic oxidation peak and may not result in any actual loss of passivity and localized corrosion. The specifications for the WP material must include allowable values for Ecorr and Ecritical. Acceptance of a material requires that (1) the measured value of Ecorr in a particular environment cannot exceed the value calculated with the corresponding correlation in Table 5 by more than 75 mV, and (2) the measured value of Ecritical in a particular environment cannot be less than the value calculated with the corresponding correlation in Table 5 by more than 75 mV. ANL-EBS-MD-000003 REV 00 49 January 2000 The correlations given in Table 5 were used to calculate the values at 10°C intervals of Ecorr and Ecritical shown in Table 6 for SDW, SCW, and SAW. The correlation for Ecritical in SSW was not used since it is based upon relatively few data points and indicates that the threshold increases with temperature, which is counter intuitive. A constant bounding value of 150 mV is assumed in this case. Table 6 shows the difference between Ecritical and Ecorr (column heading Diff.) and is never less than 150 mV between 20 and 150°C. Therefore, implementation of the potentialbased specification will prevent the use of heats of material that would be prone to passive film instability or localized corrosion. The cost of such performance would be associated with the quantity of rejected material (assumed to be approximately 20%). The specification can be changed to allow more material to be accepted but with greater risk of localized corrosion. There are precedents for using electrochemical measurements as the basis of water chemistry and materials specifications in the nuclear industry. For example, measurements of corrosion potential are indicative of dissolved oxygen and can be used to assure adequate de-aeration in various regions of the steam cycle. The role of electrochemical potential on SCC has been well documented by Andresen (1987). The critical potentials are specified as threshold potential 1 or 2. However, it must be emphasized that localized corrosion may not occur, even if these potential levels are reached. It is doubtful that localized corrosion will occur in any of these solutions, at any potential above Ecorr and below the thermodynamic limit of water. Long-term potential control experiments should be performed to determine actual values of Ecritical for Alloy 22. Clearly, more work needs to be done. In an ideal case, the crevice corrosion temperature can be estimated from the intersection of the lines representing the corrosion and threshold potentials at elevated temperature. Better correlations of Ecorr and Ecritical with material history, water chemistry, and temperature may ultimately allow precise prediction of the crevice corrosion temperature. Improved correlations would provide rigorous statistical estimates of uncertainty and variability in Ecorr and Ecritical. The precise determination of uncertainty and variability in Ecorr and Ecritical would enable designers to determine the impact of accepting 100% of the supplied WP material on repository performance. In the mean time, crevice corrosion can be forced to occur in the model by equating Ecorr and Ecritical over temperature ranges of uncertainty (90-120°C). This assumption would provide a conservative estimate of the crevice corrosion temperature. Improved LC models with accurate temperature dependence will allow a precise sensitivity study, assessing the impact of various WP design changes on the radiological dose at the site boundary. Additional work and data is needed to fill this void. ANL-EBS-MD-000003 REV 00 50 January 2000 Table 4. Compilation of Electrochemical Potentials Determined from CP Curves Sample ID Medium Temp. Reversal Potential Corrosion Potential Threshold Potential 1 Threshold Potential 2 Repassivation Potential 1 Repassivation Potential 2 Repassivation Potential 3 CP Curve Type °C mV mV mV mV mV mV mV DEA025 SDW 30 1200 -55 466 688 524 577 619 Type 1-2 DEA026 SDW 60 1200 -137 317 874 438 495 506 Type 2 DEA027 SDW 90 1200 -191 192 757 283 338 387 Type 2 DEA023 SDW 30 1200 -65 436 900 511 555 564 Type 2 DEA022 SDW 60 1200 -174 282 800 464 508 501 Type 2 DEA024 SDW 90 1190 -162 185 739 270 308 422 Type 2 DEA019 SDW 30 1190 -93 420 900 516 556 579 Type 2 DEA021 SDW 60 1190 -161 290 809 445 491 509 Type 2 DEA020 SDW 90 1200 -158 169 724 268 335 390392 Type 2 DEA009 SCW 30 795 -57 169 421 none none none Type 2 DEA011 SCW 60 797 -240 234 777 292 319 680 Type 2 DEA010 SCW 90 798 -136 206 719 16 46 663 Type 2 DEA012 SCW 30 1200 -173 338 910 490 530 699 Type 2 DEA014 SCW 60 1190 -231 226 771 319 344 572 Type 2 DEA013 SCW 90 1190 -173 336 910 490 532 699 Type 2 DEA015 SCW 30 1190 -188 341 907 538 572 742 Type 2 DEA017 SCW 60 1200 -226 238 789 323 353 595 Type 2 DEA016 SCW 90 1190 -237 199 704 609 609 622 Type 2 DTN: LL990610205924.075 ANL-EBS-MD-000003 REV 00 51 January 2000 Table 4. Compilation of Electrochemical Potentials Determined from CP Curves (Continued) Sample ID Medium Temp. Reversal Potential Corrosion Potential Threshold Potential 1 Threshold Potential 2 Repassivation Potential 1 Repassivation Potential 2 Repassivation Potential 3 CP Curve Type °C mV mV mV mV mV mV mV DEA007 SAW 30 1020 -42 663 750 none none none Type 1 DEA004 SAW 60 1050 -118 575 774 none none none Type 1 DEA002 SAW 90 1040 -176 555 642 none none none Type 1 DEA003 SAW 30 1100 -66 650 775 none none none Type 1 DEA006 SAW 60 1040 -115 613 783 none none none Type 1 DEA029 SAW 90 1200 -171 595 652 646 671 849 Type 1 DEA005 SAW 30 1820 -84 664 867 none none none Type 1 DEA008 SAW 60 1070 -102 605 708 none none none Type 1 DEA031 SAW 90 1200 -150 600 650 none none none Type 1 DEA032 SSW 100 1200 -234 234 768 none none none Type 2 DEA033 SSW 120 1200 -253 664 715 none none none Type 1 DEA035 SSW 100 1200 216 526 none none none Type 1 DEA034 SSW 120 1200 -320 171 471 none none none Type 2 DTN LL990610205924.075 Note: The term “none” indicates no detected breakdown in passive film up to the specified reversal potential; no determination of repassivation potential was possible. All potentials were measured with an Ag/AgCl reference electrode. One should subtract 197 mV from measured values to convert to the Normal Hydrogen Electrode potential scale. ANL-EBS-MD-000003 REV 00 52 January 2000 y = -4.3111x + 565 R2 = 0.9758 y = -1.6556x - 33.556 R2 = 0.7544 y = -1.4889x + 888.33 R2 = 0.2421 y = -3.0231x + 680.92 R2 = 0.943 y = -3.9278x + 698.22 R2 = 0.9259 y = -4.0556x + 656.56 R2 = 0.9334 -400 -200 0 200 400 600 800 1000 20 30 40 50 60 70 80 90 100 x = Temperature (Centigrade) y = Potential (mV vs. Ag/AgCl) Corrosion Potential Threshold Potential 1 Threshold Potential 2 Repassivation Potential 1 Repassivation Potential 2 Repassivation Potential 3 Linear (Threshold Potential 1) Linear (Corrosion Potential) Linear (Threshold Potential 2) Linear (Repassivation Potential 3) Linear (Repassivation Potential 2) Linear (Repassivation Potential 1) DTN: LL990610205924.075 Figure 12. Potentials Versus Temperature: Alloy 22 in SDW y = -0.875x + 282.93 R2 = 0.156 y = -1.0667x - 123.86 R2 = 0.1449 y = 0.7917x + 679.36 R2 = 0.0167 y = -1.2078x + 724.18 R2 = 0.1926 y = -3.4216x + 596.24 R2 = 0.1447 y = -3.0882x + 550.24 R2 = 0.111 -400 -200 0 200 400 600 800 1000 20 30 40 50 60 70 80 90 100 x = Temperature (Centigrade) y = Potential (mV vs. Ag/AgCl) Corrosion Potential Threshold Potential 1 Threshold Potential 2 Repassivation Potential 1 Repassivation Potential 2 Repassivation Potential 3 Linear (Threshold Potential 1) Linear (Corrosion Potential) Linear (Threshold Potential 2) Linear (Repassivation Potential 3) Linear (Repassivation Potential 2) Linear (Repassivation Potential 1) DTN: LL990610205924.075 Figure 13. Potentials Versus Temperature: Alloy 22 in SCW ANL-EBS-MD-000003 REV 00 53 January 2000 y = -1.1153x + 677.33 R2 = 0.5617 y = -1.9958x + 12 R2 = 0.9522 y = -2.1667x + 857 R2 = 0.641 -400 -200 0 200 400 600 800 1000 20 30 40 50 60 70 80 90 100 x = Temperature (Centigrade) y = Potential (mV vs. Ag/AgCl) Corrosion Potential Threshold Potential 1 Threshold Potential 2 Repassivation Potential 1 Repassivation Potential 2 Repassivation Potential 3 Linear (Threshold Potential 1) Linear (Corrosion Potential) Linear (Threshold Potential 2) DTN: LL990610205924.075 Figure 14. Potentials Versus Temperature: Alloy 22 in SAW y = 9.175x - 683.5 R2 = 0.1559 y = -8.75x + 1643 R2 = 0.4068 y = -2.625x + 28.5 R2 = 0.4501 -400 -200 0 200 400 600 800 1000 95 100 105 110 115 120 125 x = Temperature (Centigrade) y = Potential (mV vs. Ag/AgCl) Corrosion Potential Threshold Potential 1 Threshold Potential 2 Linear (Threshold Potential 1) Linear (Threshold Potential 2) Linear (Corrosion Potential) DTN: LL990610205924.075 Figure 15. Potentials Versus Temperature: Alloy 22 in SSW ANL-EBS-MD-000003 REV 00 54 January 2000 Table 5. Summary of Correlated Corrosion and Threshold Potential Data, Alloy 22 Figure Medium Curve Potential Parameter b0 b1 R2 12 SDW Type 2 Corrosion Ecorr -33.556 -1.6556 0.7544 12 SDW Type 2 Threshold 1 565 -4.3111 0.9758 12 SDW Type 2 Threshold 2 Ecritical 888.33 -1.4889 0.2421 12 SDW Type 2 Repassivation 1 656.56 -4.0556 0.9334 12 SDW Type 2 Repassivation 2 698.22 -3.9278 0.9259 12 SDW Type 2 Repassivation 3 680.92 -3.0231 0.943 13 SCW Type 2 Corrosion Ecorr -123.86 -1.0667 0.1449 13 SCW Type 2 Threshold 1 282.93 -0.875 0.156 13 SCW Type 2 Threshold 2 Ecritical 679.36 0.7917 0.0167 13 SCW Type 2 Repassivation 1 550.24 -3.0882 0.111 13 SCW Type 2 Repassivation 2 596.24 -3.4216 0.1447 13 SCW Type 2 Repassivation 3 724.18 -1.2078 0.1926 14 SAW Type 1 Corrosion Ecorr 12 -1.9958 0.9522 14 SAW Type 1 Threshold 1a Ecritical 677.33 -1.1153 0.5617 14 SAW Type 1 Threshold 1b 857 -2.1667 0.641 15 SSW Type 1 Corrosion Ecorr 28.5 -2.625 0.4501 15 SSW Type 1 Threshold 1a Ecritical -683.5 9.175 0.1559 15 SSW Type 1 Threshold 1b 1643 -8.75 0.4068 NOTE: R2 is the regression coefficient. Table 6. Values of Ecorr and Ecritical Based on Correlated CP Data, Alloy 22 SDW SDW SDW SCW SCW SCW SAW SAW SAW SSW SSW SSW T Ecorr Ecritical Diff. Ecorr Ecritical Diff. Ecorr Ecritical Diff. Ecorr Ecritical Diff. (°C) (mV) (mV) (mV) (mV) (mV) (mV) (mV) (mV) (mV) (mV) (mV) (mV) 20 -67 479 545 -145 265 411 -28 655 683 -24 150 174 30 -83 436 519 -156 257 413 -48 644 692 -50 150 200 40 -100 393 492 -167 248 414 -68 633 701 -77 150 227 50 -116 349 466 -177 239 416 -88 622 709 -103 150 253 60 -133 306 439 -188 230 418 -108 610 718 -129 150 279 70 -149 263 413 -199 222 420 -128 599 727 -155 150 305 80 -166 220 386 -209 213 422 -148 588 736 -182 150 332 90 -183 177 360 -220 204 424 -168 577 745 -208 150 358 100 -199 134 333 -231 195 426 -188 566 753 -234 150 384 110 -216 91 306 -241 187 428 -208 555 762 -260 150 410 120 -232 48 280 -252 178 430 -227 543 771 -287 150 437 130 -249 5 253 -263 169 432 -247 532 780 -313 150 463 140 -265 -39 227 -273 160 434 -267 521 789 -339 150 489 150 -282 -82 200 -284 152 436 -287 510 797 -365 150 515 ANL-EBS-MD-000003 REV 00 55 January 2000 6.4.4 Effect of Gamma Radiolysis on Corrosion Potential Anodic shifts in the open circuit corrosion potential of stainless steel have been experimentally observed (Glass et al. 1986; Kim 1987, 1988, 1999a, 1999b). Glass et al. (1986) performed ambient-temperature CP of 316L samples in 0.018 M NaCl solution during exposure to 3.5 Mrad h-1 gamma radiation. He found that the corrosion current shifted in the anodic direction by approximately 200 mV. From inspection of the graphical data in this article, it is concluded that there is very little increase in the corresponding corrosion current density. However, the separation between the corrosion potential and the threshold for localized attack decreased slightly. This shift in corrosion potential was shown to be due to the formation of hydrogen peroxide. This finding was subsequently confirmed by Kim (1988). In this case, ambienttemperature CP of 316 stainless steel in acidic (pH~2) 1.5 M NaCl during exposure to 0.15 Mrad h-1 gamma radiation showed a 100 mV anodic shift in the corrosion potential, with very little effect on the corrosion current. Note that Glass et al. (1986) and Kim (1988) worked on stainless steels, not Alloy 22. Additional studies of the corrosion and threshold potentials of Alloy 22 in the presence of gamma radiation, as done by Glass et al. in the early 1980’s, is beyond the YMP’s current work scope. To determine the maximum impact that gamma radiolysis could have on the corrosion potential, hydrogen peroxide was added to electrolytes used for testing Alloy 22. Experiments at 25°C are illustrated by Figures 16 and 17. As the concentration of hydrogen peroxide in SAW approaches 72 ppm (calculated from number of added drops of H2O2), the corrosion potential asymptotically approaches 150 mV versus Ag/AgCl, well below any threshold where localized attach would be expected in SAW. Similarly, as the concentration of hydrogen peroxide in SCW approaches 72 ppm, the corrosion potential asymptotically approaches -25 mV versus Ag/AgCl, well below any threshold where localized attach would be expected in SCW. This change in corrosion potential is also below any level where a change in oxidation state would be expected. Since extremely high radiation levels would be required to achieve such shifts in corrosion potential and since even the maximum shifts in potential would be less than those required for breakdown of the passive film, it seems unlikely that gamma radiolysis will lead to catastrophic failure of Alloy 22 due to LC. However, as more resources become available, actual tests with a gamma source should be performed. ANL-EBS-MD-000003 REV 00 56 January 2000 -100 -50 0 50 100 150 200 0 1 2 3 4 5 6 Time (103 Seconds) Potential (mV vs. Ag/AgCl) Addition of 25 microliters of 30% H2O2 (~ 8 ppm) No H2O2 (~ 0 ppm) 0 8 16 24 32 40 48 56 64 72 Figure 16. Effect of Hydrogen Peroxide on Corrosion Potential of Alloy 22 in SAW at 25°C -250 -200 -150 -100 -50 0 0 1 2 3 4 5 6 Time (103 seconds) Potential (mV vs. Ag/AgCl) Addition of 25 microliters of 30% H2O2 (~ 8 ppm) No H2O2 (~ 0 ppm) 0 8 16 24 32 40 48 56 64 72 Figure 17. Effect of Hydrogen Peroxide on Corrosion Potential of Alloy 22 in SCW at 25°C ANL-EBS-MD-000003 REV 00 57 January 2000 6.4.5 Correction of Measured Potential for Junction Potential It is important to understand the magnitude of the error in the potential measurements due to the junction potential. A correction has been performed based upon the Henderson Equation (Bard and Faulkner 1980). ( ) ( ) [ ] ( ) ( ) [ ] ( ) ( ) ß a a ß a ß i i i i i i i i i i i i i i i i i i i j C u z C u z F RT C C u z C C z u z E . . . . - - = ln (Eq. 18) where Ej is the potential across the junction connecting the a and ß phases, zi is the valence of the ith ion, ui is the mobility of the ith ion, Ci( a)is the concentration of the ith ion in the a phase, Ci( ß) is the concentration of the ith ion in the ß phase, R is the universal gas constant, T is the absolute temperature, and F is Faraday’s constant. Calculated values of Ej for the isothermal junction are summarized in Table 7 and required the summation of various products such as ( ) a i i i C u z , ( ) ß i i i C u z , ( ) ( ) [ ] a ß i i i i C C u z - and ( ) ( ) [ ] i i i i i z C C u z a ß - . Table 7. Summary of Junction Potential Corrections for CP (volts) T (°C) SDW SCW SAW SSW 30 2.716E-03 1.188E-03 6.019E-03 -7.649E-03 60 2.984E-03 1.306E-03 6.615E-03 -8.406E-03 90 3.253E-03 1.423E-03 7.210E-03 -9.164E-03 A positive value indicates that the electrochemical potential on the KCl side of the junction (ß phase) is greater than the electrochemical potential in the test medium (a phase), in close proximity to the Luggin probe. The potential in the test medium can be calculated from the measured value by subtracting Ej. The calculated junction potentials in Table 7 are supported by the data in Tables 8 through 11. Ionic properties used in the calculation were taken from Bard and Faulkner (1980). These corrections are not very large, with the largest being less than 9 mV. This value corresponds to the junction potential for SSW at 90°C. It is concluded that insignificant error results from neglecting to correct for the junction potential. ANL-EBS-MD-000003 REV 00 58 January 2000 Table 8. Junction Potential Correction for CP with SDW i FW ( ) a i C i z i z i u ( ) a i i i C u z ( ) ß i i i C u z ( ) ( ) [ ] a ß i i i i C C u z - ( ) ( ) [ ] i i i i i z C C u z a ß - (mol/L-1) (cm2 sec-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) K+1 39.0983 8.696E-04 1 1 7.62E-04 6.626E-07 3.048E-03 3.047E-03 3.047E-03 Na+1 22.9898 1.779E-02 1 1 5.19E-04 9.239E-06 0 -9.239E-06 -9.239E-06 Mg+2 24.3050 4.114E-05 2 2 5.00E-04 4.114E-08 0 -4.114E-08 -2.057E-08 Ca+2 40.0780 1.248E-05 2 2 6.17E-04 1.538E-08 0 -1.538E-08 -7.692E-09 F-1 18.9984 7.369E-04 -1 1 5.00E-04 3.685E-07 0 -3.685E-07 3.685E-07 Cl-1 35.4527 1.890E-03 -1 1 7.91E-04 1.495E-06 3.165E-03 3.163E-03 -3.163E-03 NO3 -1 62.0049 1.032E-03 -1 1 7.40E-04 7.642E-07 0 -7.642E-07 7.642E-07 SO4 -2 96.0636 1.738E-03 -2 2 8.27E-04 2.875E-06 0 -2.875E-06 1.438E-06 HCO3 -1 61.0171 1.552E-02 -1 1 4.61E-04 7.155E-06 0 -7.155E-06 7.155E-06 SiO3 -2 76.0837 9.614E-04 -2 2 5.00E-04 9.614E-07 0 -9.614E-07 4.807E-07 H+1 1.0079 7.943E-09 1 1 3.63E-03 2.879E-11 0 -2.879E-11 -2.879E-11 pH 8.100E+00 Summation 2.358E-05 6.212E-03 6.189E-03 -1.154E-04 j E 2.716E-03 Volts at 30°C Beta - Alpha j E 2.984E-03 Volts at 60°C Beta - Alpha j E 3.253E-03 Volts at 90°C Beta - Alpha ANL-EBS-MD-000003 REV 00 59 January 2000 Table 9. Junction Potential Correction for CP with SCW i FW ( ) a i C i z i z i u ( ) a i i i C u z ( ) ß i i i C u z ( ) ( ) [ ] a ß i i i i C C u z - ( ) ( ) [ ] i i i i i z C C u z a ß - mol/L-1 (cm2 sec-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) K+1 39.0983 8.696E-02 1 1 7.62E-04 6.626E-05 3.048E-03 2.981E-03 2.981E-03 Na+1 22.9898 1.779E+00 1 1 5.19E-04 9.239E-04 0 -9.239E-04 -9.239E-04 Mg+2 24.3050 4.114E-05 2 2 5.00E-04 4.114E-08 0 -4.114E-08 -2.057E-08 Ca+2 40.0780 2.495E-05 2 2 6.17E-04 3.077E-08 0 -3.077E-08 -1.538E-08 F-1 18.9984 7.369E-02 -1 1 5.00E-04 3.685E-05 0 -3.685E-05 3.685E-05 Cl-1 35.4527 1.890E-01 -1 1 7.91E-04 1.495E-04 3.165E-03 3.015E-03 -3.015E-03 NO3 -1 62.0049 1.032E-01 -1 1 7.40E-04 7.642E-05 0 -7.642E-05 7.642E-05 SO4 -2 96.0636 1.738E-01 -2 2 8.27E-04 2.875E-04 0 -2.875E-04 1.438E-04 HCO3 -1 61.0171 1.147E+00 -1 1 4.61E-04 5.289E-04 0 -5.289E-04 5.289E-04 SiO3 -2 76.0837 9.614E-04 -2 2 5.00E-04 9.614E-07 0 -9.614E-07 4.807E-07 H+1 1.0079 7.943E-09 1 1 3.63E-03 2.879E-11 0 -2.879E-11 -2.879E-11 pH 8.100E+00 Summation 2.070E-03 6.212E-03 4.142E-03 -1.714E-04 j E 1.188E-03 Volts at 30°C Beta - Alpha j E 1.306E-03 Volts at 60°C Beta - Alpha j E 1.423E-03 Volts at 90°C Beta - Alpha ANL-EBS-MD-000003 REV 00 60 January 2000 Table 10. Junction Potential Correction for CP with SAW i FW ( ) a i C i z i z i u ( ) a i i i C u z ( ) ß i i i C u z ( ) ( ) [ ] a ß i i i i C C u z - ( ) ( ) [ ] i i i i i z C C u z a ß - mol L-1 (cm2 sec-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) K+1 39.0983 8.696E-02 1 1 7.62E-04 6.626E-05 3.048E-03 2.981E-03 2.981E-03 Na+1 22.9898 1.779E+00 1 1 5.19E-04 9.239E-04 0 -9.239E-04 -9.239E-04 Mg+2 24.3050 4.114E-02 2 2 5.00E-04 4.114E-05 0 -4.114E-05 -2.057E-05 Ca+2 40.0780 2.495E-02 2 2 6.17E-04 3.077E-05 0 -3.077E-05 -1.538E-05 F-1 18.9984 0.000E+00 -1 1 5.00E-04 0.000E+00 0 0.000E+00 0.000E+00 Cl-1 35.4527 1.890E-01 -1 1 7.91E-04 1.495E-04 3.165E-03 3.015E-03 -3.015E-03 NO3 -1 62.0049 1.032E-01 -1 1 7.40E-04 7.642E-05 0 -7.642E-05 7.642E-05 SO4 -2 96.0636 1.738E-01 -2 2 8.27E-04 2.875E-04 0 -2.875E-04 1.438E-04 HCO3 -1 61.0171 0.000E+00 -1 1 4.61E-04 0.000E+00 0 0.000E+00 0.000E+00 SiO3 -2 76.0837 9.614E-04 -2 2 5.00E-04 9.614E-07 0 -9.614E-07 4.807E-07 H+1 1.0079 1.995E-03 1 1 3.63E-03 7.233E-06 0 -7.233E-06 -7.233E-06 pH 2.700E+00 Summation 1.584E-03 6.212E-03 4.629E-03 -7.803E-04 j E 6.019E-03 Volts at 30°C Beta - Alpha j E 6.615E-03 Volts at 60°C Beta - Alpha j E 7.210E-03 Volts at 90°C Beta - Alpha ANL-EBS-MD-000003 REV 00 61 January 2000 Table 11. Junction Potential Correction for CP with SSW i FW ( ) a i C i z i z i u ( ) a i i i C u z ( ) ß i i i C u z ( ) ( ) [ ] a ß i i i i C C u z - ( ) ( ) [ ] i i i i i z C C u z a ß - mol L-1 (cm2 sec-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) (mol cm-1 s-1 V-1) K+1 39.0983 3.622E+00 1 1 7.62E-04 2.759E-03 3.048E-03 2.882E-04 2.882E-04 Na+1 22.9898 2.118E+01 1 1 5.19E-04 1.100E-02 0 -1.100E-02 -1.100E-02 Mg+2 24.3050 0.000E+00 2 2 5.00E-04 0.000E+00 0 0.000E+00 0.000E+00 Ca+2 40.0780 0.000E+00 2 2 6.17E-04 0.000E+00 0 0.000E+00 0.000E+00 F-1 18.9984 0.000E+00 -1 1 5.00E-04 0.000E+00 0 0.000E+00 0.000E+00 Cl-1 35.4527 3.622E+00 -1 1 7.91E-04 2.865E-03 3.165E-03 2.993E-04 -2.993E-04 NO3 -1 62.0049 2.113E+01 -1 1 7.40E-04 1.564E-02 0 -1.564E-02 1.564E-02 SO4 -2 96.0636 0.000E+00 -2 2 8.27E-04 0.000E+00 0 0.000E+00 0.000E+00 HCO3 -1 61.0171 0.000E+00 -1 1 4.61E-04 0.000E+00 0 0.000E+00 0.000E+00 SiO3 -2 76.0837 0.000E+00 -2 2 5.00E-04 0.000E+00 0 0.000E+00 0.000E+00 H+1 1.0079 1.000E-07 1 1 3.63E-03 3.625E-10 0 -3.625E-10 -3.625E-10 pH 7.000E+00 Summation 3.227E-02 6.212E-03 -2.606E-02 4.631E-03 j E -7.649E-03 Volts at 30°C Beta - Alpha j E -8.406E-03 Volts at 60°C Beta - Alpha j E -9.164E-03 Volts at 90°C Beta - Alpha ANL-EBS-MD-000003 REV 00 62 January 2000 6.5 RATES OF GENERAL AQUEOUS-PHASE CORROSION GC rates are assumed if the threshold potential (Ecritical) is not exceeded. GC rates have been estimated with weight-loss data from the LTCTF (Estill 1998). LC rates and failure mode characteristics (e.g., number failure sites and opening size) will have to be estimated from other published data. Only estimates of LC rates are given in this report. Since pitting has not been observed in LTCF experiments at LLNL, it is assumed that the primary mode of LC is crevice corrosion. This aqueous phase general and localized corrosion model will be applied to each element (patch) in the WAPDEG simulation. To the extent possible, uncertainty will be estimated from available data. 6.5.1 Corrosion Rates Based Upon Electrochemical Measurements The corrosion (or penetration) rate of an alloy can be calculated from the corrosion current density with the following formula derived from Jones (1996): F n i dt dp alloy alloy corr . = (Eq. 19) where p is the penetration depth, t is time, icorr is the corrosion current density, .alloy is the density of the alloy, assumed to be approximately 8.69 g cm-3 for Alloy 22, nalloy, is the number of gram equivalents per gram of alloy, and F is Faraday’s constant. The value of nalloy can be calculated with the following formula: . . . . . . . . . = j j j j alloy a n f n (Eq. 20) where fj is the mass fraction of the jth alloying element in the material, nj is the number of electrons involved in the anodic dissolution process, which is assumed to be congruent, and aj is the atomic weight of the jth alloying element. Congruent dissolution means that the dissolution rate of a given alloy element is proportional to its concentration in the bulk alloy. These equations have been used to calculate the penetration rate for Alloy 22 as a function of corrosion current density. The results of these calculations are shown in Tables 12 and 13. Values of (fjnj/aj)/100 must be summed to calculate dp/dt. While calculated values of dp/dt are believed to have only three significant figures, values of (fjnj/aj)/100 are given with more figures. By carrying the extra figures during calculation (and checking), round-off errors in final results can be minimized. In subsequent versions of the AMR, fewer significant figures will be reported. The penetration rate for Alloy 22 is linearly proportional to current density and is estimated to be between 9.39 and 9.73 µm per year at 1 µA cm-2. Usually, the corrosion current density, icorr, is determined from the intersection of the anodic and cathodic Tafel lines at Ecorr (Jones 1996; Bard and Faulkner 1980). However, this assumes that Butler-Volmer kinetics apply at the interface. Since the Alloy 22 surface is passivated with a protective oxide film, this may not be true. In fact, the cathodic curves from Ecorr have limited Tafel linearity in Figures 3 through 9. Nevertheless, approximate Tafel extrapolations generally yield icorr values around 1×10-6 A cm-2, which are about one hundred times higher than the ANL-EBS-MD-000003 REV 00 63 January 2000 equivalent icorr from LTCTF weight loss data. Tafel extrapolations should give much lower icorr when the specimen electrodes are pre-exposed for times much greater than the one hour specified by ASTM G 5-87 (ASTM 1989) because the passive corrosion rate decreases logarithmically with time. Given these non-idealities, the local minima in current observed at Ecorr (circled in Figure 11) has been interpreted as a lower bound for the corrosion current density, icorr, not as the corrosion current density per se. The non-equilibrium passive current density, ipass, serves as the upper bound of the corrosion current density. It is believed that the local minima (circled) are relatively close to the corrosion current density, the point at which the anodic and cathodic processes are balanced. Note that current (A) and current density (A cm-2) are practically the same since the exposed area of the sample is about one square centimeter (0.96 cm2). In principle, electrochemically determined rates should be consistent with those observed in the LTCTF. To a first order approximation, such consistency appears to exist between most of the circled current densities (lower bound of the corrosion current densities) and the LTCTF results. GC rates from the LTCTF appear to be normally distributed around a mean value. The median GC rate based upon all Alloy 22 weight loss samples is approximately 16.51 nm y-1 (0.01651 µm per year). See Section 6.5.3. Assuming a penetration rate of 9.73 µm y-1 at a corrosion current density of 1 µA cm-2, the median corrosion rate in the LTCTF corresponds to an apparent corrosion current density of approximately 1.70×10-9 A cm-2. As can be seen in Figure 11, this value lies within the range of observed lower bounds of the density, which covers a range extending from 6×10-10 to 2×10-8 A cm-2. Since the instrument appeared to have difficulty measuring extremely low current levels, values of 10-13 A cm-2 are ignored. The lower bounds of the corrosion currents and the non-equilibrium passive currents from CP measurements in SDW, SCW, SAW, and SSW are summarized in Figures 18 through 21. In general, it has been found that the current verses temperature data can be represented by the following linear regression equation: x b b y ln ln ln 1 0 + = (Eq. 21) where y is the current (A) and x is the temperature (°C). This can be rewritten as follows: ( )1 0 b x b y × = (Eq. 22) Since the exposed area in these measurements is approximately 0.96 cm2, the current density is approximately equal to the current. The coefficients based upon the correlation of data for SDW, SCW, SAW, and SSW are summarized in Table 14. These coefficients were used to calculate the bounding values given in Table 15. The ranges of current density are converted to ranges of corrosion rate based upon the information in Tables 12 and 13. ANL-EBS-MD-000003 REV 00 64 January 2000 Table 12. Conversion of Current Density to Corrosion (Penetration) Rate – Result Units Value at Low fi Value at High fi Faraday Constant C equiv-1 9.648460E+04 9.648460E+04 Assumed Current Density A cm-2 1.000000E-06 1.000000E-06 Assumed Mass Density g cm-3 8.690000E+00 8.690000E+00 Total (fjnj/aj)/100 3.864793E-02 4.005512E-02 dp/dt cm sec-1 3.086000E-11 2.977585E-11 dp/dt µm per year 9.732010E+00 9.390113E+00 Table 13. Conversion of Current Density to Corrosion (Penetration) Rate – Calculation aj nj nj nj fj fj (fjnj/aj)/100 (fjnj/aj)/100 wt% wt% Low High Calc. Low High Low High C 12.011 2 4 4 0.015 0.015 4.995421E-05 4.995421E-05 Ni 58.69 2 3 2 62.365 49.535 2.125234E-02 1.688022E-02 Cr 51.9969 3 6 3 20.000 22.500 1.153915E-02 1.298154E-02 Mo 95.94 3 6 3 12.500 14.500 3.908693E-03 4.534084E-03 Fe 55.847 2 3 2 2.000 6.000 7.162426E-04 2.148728E-03 Cu 63.546 1 2 2 0.000 0.000 0.000000E+00 0.000000E+00 P 30.973762 3 5 5 0.020 0.020 3.228539E-05 3.228539E-05 Si 28.0855 4 4 4 0.080 0.080 1.139378E-04 1.139378E-04 S 32.066 2 6 6 0.020 0.500 3.742282E-05 9.355704E-04 Mn 54.93805 2 2 2 0.500 0.500 1.820232E-04 1.820232E-04 W 183.85 2 6 6 2.500 3.500 8.158825E-04 1.142236E-03 Co 58.9332 2 3 2 0.000 2.500 0.000000E+00 8.484182E-04 V 50.9415 2 3 3 0.000 0.350 0.000000E+00 2.061188E-04 Ti 47.88 2 3 3 0.000 0.000 0.000000E+00 0.000000E+00 Pd 105.42 2 2 2 0.000 0.000 0.000000E+00 0.000000E+00 Other 1 0 0 0 0.000 0.000 0.000000E+00 0.000000E+00 Total 100.000 100.000 3.864793E-02 4.005512E-02 ANL-EBS-MD-000003 REV 00 65 January 2000 y = (1E-16)x3.7402 R2 = 0.8288 y = (2E-05)x-0.5453 R2 = 0.2869 1.0E-11 1.0E-10 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 20 30 40 50 60 70 80 90 100 x = Temperature (Centigrade) y = Current (amps) Lower Bound of Corrosion Current Non-Equilibrium Passive Current Power (Lower Bound of Corrosion Current) Power (Non-Equilibrium Passive Current) DTN: LL990610205924.075 Figure 18. Bounding Currents Versus Temperature, Alloy 22 in SDW y = (6E-07)x0.6935 R2 = 0.2258 y = (1E-30)x11.866 R2 = 0.853 1.0E-11 1.0E-10 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 20 30 40 50 60 70 80 90 100 x = Temperature (Centigrade) y = Current (amps) Lower Bound of Corrosion Current Non-Equilibrium Passive Current Power (Non-Equilibrium Passive Current) Power (Lower Bound of Corrosion Current) DTN: LL990610205924.075 Figure 19. Bounding Currents Versus Temperature, Alloy 22 in SCW ANL-EBS-MD-000003 REV 00 66 January 2000 y = (5E-07)x0.549 R2 = 0.5907 y = (2E-19)x5.5868 R2 = 0.6506 1.0E-11 1.0E-10 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 20 30 40 50 60 70 80 90 100 x = Temperature (Centigrade) y = Current (amps) Lower Bound of Corrosion Current Non-Equilibrium Passive Current Power (Non-Equilibrium Passive Current) Power (Lower Bound of Corrosion Current) DTN: LL990610205924.075 Figure 20. Bounding Currents Versus Temperature, Alloy 22 in SAW y = (2E-17)x5.6236 R2 = 0.2341 1.00E-11 1.00E-10 1.00E-09 1.00E-08 1.00E-07 1.00E-06 1.00E-05 1.00E-04 95 100 105 110 115 120 125 x = Temperature (Centigrade) y = Current (amps) Non-Equilibrium Passive Current Power (Non-Equilibrium Passive Current) DTN: LL990610205924.075 Figure 21. Bounding Currents Versus Temperature, Alloy 22 in SSW ANL-EBS-MD-000003 REV 00 67 January 2000 Table 14. Coefficients for Regression Equations Used to Represent Lower Bounds of Corrosion Current and Non-Equilibrium Passive Current Figure Medium Curve Current Corresponding Potential b0 b1 R2 18 SDW Type 2 Corrosion Ecorr 1.0E-16 3.7402 0.8288 18 SDW Type 2 Passive Ecritical 2.0E-05 -0.5453 0.2869 19 SCW Type 2 Corrosion Ecorr 1.0E-30 11.866 0.853 19 SCW Type 2 Passive Ecritical 6.0E-07 0.6935 0.2258 20 SAW Type 1 Corrosion Ecorr 2.0E-19 5.5868 0.6506 20 SAW Type 1 Passive Ecritical 5.0E-07 0.549 0.5907 21 SSW Type 1 Corrosion Ecorr 21 SSW Type 1 Passive Ecritical 2.0E-17 5.6236 0.2341 DTN: LL990610205924.075 Table 15. Rates Based Upon of Correlations of Lower Bounds of Corrosion Current and Non-Equilibrium Passive Current Figure Medium Curve Basis of Estimate Current – Temp. Current (µA) Current Density (µA cm-2) Corrosion Rate (µm y-1) 18 SDW Type 2 Corrosion – 90°C 2.04E-03 2.12E-03 1.99E-02 18 SDW Type 2 Passive – 90°C 1.72 1.79 16.8 19 SCW Type 2 Corrosion – 90°C 0.155 0.161 1.51 19 SCW Type 2 Passive – 90°C 13.6 14.2 133.0 20 SAW Type 1 Corrosion – 90°C 1.66E-02 1.73E-02 0.162 20 SAW Type 1 Passive – 90°C 1.18 1.23 11.6 21 SSW Type 1 Corrosion – 90°C 21 SSW Type 1 Passive – 90°C 9.82 10.3 96.4 DTN: LL990610205924.075 6.5.2 Corrosion Rates Based Upon Weight Loss Measurements The LTCTF provides a complete source of corrosion data for Alloy 22 in environments relevant to the proposed high-level waste repository at Yucca Mountain. The LTCTF and results from that facility are described in detail in previous publications by the YMP (Estill 1998). The GC rates of Alloy 22 measured in the LTCTF should be representative of those expected in the repository. Testing includes a wide range of plausible generic test media, including SDW, SCW, Simulated Cement-Modified Water, and SAW. The SCW test medium is three orders-ofmagnitude (1000×) more concentrated than J-13 well water and is slightly alkaline (pH~8). The SAW test medium is three orders-of-magnitude (1000×) more concentrated than J-13 well water and is acidic (pH~2.7) to mimic the evaporative concentration of electrolytes on the hot WP surface. Concentrated solutions are intended to mimic the evaporative concentration of the electrolytes on the hot WP surface. Two temperature levels (60 and 90°C) are included. The ANL-EBS-MD-000003 REV 00 68 January 2000 maximum observed rate, which is much less than 1 µm per year, clearly indicates that the life of the Alloy 22 outer barrier will not be limited by GC. It is also assumed that the corrosion rate is constant and does not decay with time. Less conservative corrosion models assume that the rate decays with time. This facility is equipped with an array of fiberglass tanks. Each tank has a total volume of ~2000 L and is filled with ~1000 L of aqueous test solution. The solution in a particular tank is controlled at either 60 or 90°C, covered with a blanket of air flowing at approximately 150 cm3 min-1, and agitated. The descriptions and compositions of three of these solutions are summarized in Table 3. Four generic types of samples, U-bends, crevices, weight loss samples, and galvanic couples, are mounted on insulating racks and placed in the tanks. Approximately half of the samples are submersed, half are in the saturated vapor above the aqueous phase, and a limited number are at the water line. It is important to note that condensed water is present on specimens located in the saturated vapor. After racks of samples were removed from the tank, samples were first rinsed with deionized water to remove salt solutions. Samples discussed have generic weight-loss or crevice geometry. Generic weight-loss samples were rectangular in shape (1 inch wide, 2 inches long, 1/8 inch thick). Generic crevice samples were square with a hole in the center (2 inches on each side, 1/8 inch thick, with a 0.312 inch diameter hole). Next, samples were removed from the rack by loosening fixture mounts with standard wrenches. The crevice assemblies described by Estill (1998) required further disassembly, which was also done with standard wrenches. After dismounting and disassembly, the metal samples of Alloy 22 were cleaned with the solution designated C.7.5 for stainless steels given in Table A1 of ASTM G 1-90 (ASTM 1997e). This solution consists of 100 ml HNO3 (specific gravity ~ 1.42) and 20 ml HF (specific gravity ~ 1.198) in enough water to give a total volume of 1000 ml. Note that alternative solutions for nickel and nickel-based alloys, designated C.6.1 and C.6.2, could have also been used. These cleaner formulations are based upon aqueous solutions of HCl and H2SO4, respectively. The crevice samples were configured in such a way as to reveal crevice corrosion if it occurred. Since no crevice attack was observed with the samples represented by these figures, it is assumed that all weight loss in the crevice samples was due to GC outside of the crevice region (area underneath washer). This is consistent with other ex situ examinations. As previously discussed, GC measurements are based upon ASTM G 1-81 (ASTM 1987) or the more recent ASTM G 1-90 (ASTM 1997e). The GC (or penetration) rate of an alloy can be calculated from weight loss data as follows with the following general formula: ( ) ( )D T A W K Rate Corrosion × × × = (Eq. 23) where K is a constant, T is the time of exposure in hours, A is the exposed area of the sample in square centimeters, W is the mass loss in grams, and D is the density in grams per cubic centimeter. The value of K used for the LTCTF data was 8.76×107 µm per year. This formula for corrosion rate can be rewritten in the following form: ANL-EBS-MD-000003 REV 00 69 January 2000 ( ) ( ) ( ) [ ]c a c b b a t w dt dp × + × + × × = 2 2 2 1 . (Eq. 24) where dp/dt is the corrosion rate, w is the mass loss in grams, . is the density in grams per cubic centimeter, t is the time of exposure in years, and the quantity in square brackets represents the exposed area of the sample in square centimeters. Without application of any conversion factor, the corrosion rate calculated with this formula has the units of centimeters per year. Multiplication of dp/dt by 104 µm cm-1 yields a corrosion rate with the units of µm per year. The weight loss and dimensional change were measured with electronic instruments calibrated to traceable standards. All data was digitally transferred to computer, minimizing the possibility of human typographical error. Comparative sample calculations are used to compare the two formulae. With specific values assumed for the purpose of comparison, the first formula yields: K = 8.76 ×107 µm y-1h cm-1 W = 0.0001 gm A = 1.0 cm 2 T = 4320 h D = 8.69 gm cm-3 Corrosion Rate = 8.76 × 107 µm yr-1h cm-1 ( )0.0001 gm ( ) 1.0 cm 2 ( )4320 h ( )8.69 gm cm-3 ( )= 0.23 µm y-1 The density for Alloy 22 used in this sample calculation was taken from Section 7.1 of ASTM B 575-94 (ASTM 1997a). A calculation with the second formula and the same assumed values gives an identical result: k = 104 µmcm-1 w = 0.0001 gm 2 a × b ( )+ 2 b × c ( )+ 2 a × c ( )=1.0 cm 2 t = 0.5 y .= 8.69 gm cm-3 dp dt = 104 µm cm-1 ( )0.0001 gm ( ) 1.0 cm 2 ( )0.5 yr ( )8.69 gm cm-3 ( )= 0.23 µm y-1 The second formula is used as the basis of a formal error analysis of GC rates determined from LTCTF data. All GC rates for Alloy 22 based on LTCTF weight loss samples are shown in Figure 22. It appears that these measurements are independent of temperature between 60 and 90°C. Furthermore, the composition of the test medium (SDW, SCW, or SAW) appeared to have little impact on the measurements. Since the maximum observed rate is only 160 nm y-1, it is ANL-EBS-MD-000003 REV 00 70 January 2000 concluded that the actual corrosion rate is below the detectable level. When all of the measured corrosion rates based upon the weight loss samples are ranked together, regardless of the test medium or temperature, the data appear to be normally distributed around a median value. This is illustrated by Figure 23. All GC rates for Alloy 22 based on LTCTF crevice samples are shown in Figure 24 (rates based on areas outside of crevice). In this case, it also appears that the measurements are independent of temperature and test medium. When all of the measured corrosion rates based upon the weight loss samples are ranked together as shown in Figure 25, most of the data points fall below 160 nm y-1 and appear to be normally distributed around a median value. However, there are four data points that appear to lie above the detection limit (between 200 and 750 nm, per year). Since no crevice attack of these four samples is evident with microscopic examination, it is believed that these points are due to the accidental removal of material during mechanical assembly of the crevice sample (Section 6.5.5). The largest measured rate shown in Figure 25 will not lead to failure of the WP during the 10,000 year service life. Based upon these data, it does not appear that the life of the WP will be limited to less than 10,000 years by the GC of Alloy 22 at temperatures less than those involved in the test (90°C). The mean and standard deviation are also determined through calculation (Burr 1974). The average corrosion rate based upon all weight loss samples is 20 nm y-1 with a standard deviation of 40 nm y-1. This compares reasonably well with the values obtained by inspection of the plotted data in Figure 23. The average corrosion rate based upon all crevice samples is 71 nm y-1 with a standard deviation of 89 nm y-1. If the four highest rates are omitted, the average rate is then calculated to be 57 nm y-1 with a standard deviation of 40 nm y-1. This is consistent with the plotted data in Figure 25. It should be noted that the distribution of corrosion rates includes some negative values. The negative corrosion rates correspond to cases where the samples actually appear to have gained weight during exposure, due to oxide growth or the formation of silicate deposits. To substantiate these interpretations, AFM has been used to inspect a number of samples removed from the LTCTF. Results are given in Section 6.5.6 and in Attachment I. ANL-EBS-MD-000003 REV 00 71 January 2000 -100 -50 0 50 100 150 200 50 60 70 80 90 100 Temperature (Centigrade) Rate (nm/yr) SAW (nm/yr) SCW (nm/yr) SDW (nm/yr) Linear (SAW (nm/yr)) Linear (SCW (nm/yr)) Linear (SDW (nm/yr)) DTN: LL990610605924.079 Figure 22. GC of Alloy 22, 6, and 12 Month Weight Loss Samples from LTCTF, Corrosion Rate Versus Temperature 0 10 20 30 40 50 60 70 80 90 100 -100 -50 0 50 100 150 200 Penetration Rate (nm/yr) Percentile (%) Median DTN: LL990610605924.079 Figure 23. GC of Alloy 22, 6, and 12 Month Weight Loss Samples from LTCTF, Percentile Versus Corrosion Rate ANL-EBS-MD-000003 REV 00 72 January 2000 -100 0 100 200 300 400 500 600 700 800 50 60 70 80 90 100 Temperature (Centigrade) Penetration Rate (nm/yr) SAW (nm/yr) SCW (nm/yr) SDW (nm/yr) Linear (SAW (nm/yr)) Linear (SCW (nm/yr)) Linear (SDW (nm/yr)) DTN: LL990610605924.079 Figure 24. GC of Alloy 22, 6, and 12 Month Crevice Samples from LTCTF, Corrosion Rate Versus Temperature 0 10 20 30 40 50 60 70 80 90 100 -10 0 -50 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 Penetration Rate (nm/yr) Percentile (%) Median DTN: LL990610605924.079 Figure 25. GC of Alloy 22, 6, and 12 Month Crevice Samples from LTCTF, Percentile Versus Corrosion Rate ANL-EBS-MD-000003 REV 00 73 January 2000 6.5.3 Error Analysis for Weight Loss Measurements The general method used in the formal error analysis is now presented and is important since it enables sound interpretation of the data shown in Figures 23 and 25. Consider the dependent variable y defined by the following generic function: ) , , , ( 4 3 2 1 n x x x x x f y · · · = (Eq. 25) where xi is the ith independent variable. The total derivative of y is then defined as follows: n n dx x y dx x y dx x y dx x y dx x y dy . . + · · · + . . + . . + . . + . . = 4 4 3 3 2 2 1 1 (Eq. 26) Based upon this definition, the maximum error in y can then be defined as: n n x x y x x y x x y x x y x x y y . . . + · · · + . . . + . . . + . . . + . . . = . 4 4 3 3 2 2 1 1 (Eq. 27) where .xi is the error in the ith independent variable. Let the dependent variable y be the GC rate measured in the LTCTF: ( ) ( ) ( ) [ ]c a c b b a t w dt dp y × + × + × × = = 2 2 2 1 . (Eq. 28) The total derivative of the corrosion rate is: dc c y db b y da a y dt t y d y dw w y dy . . + . . + . . + . . + . . + . . = . . (Eq. 29) The maximum error in the corrosion rate is: c c y b b y a a y t t y y w w y y . . . + . . . + . . . + . . . + . . . + . . . = . . . (Eq. 30) The partial derivatives are: ( ) ( ) ( ) [ ]c a c b b a t w y × + × + × × = . . 2 2 2 1 1 . (Eq. 31) ( ) ( ) ( ) [ ]c a c b b a t w y × + × + × × = . . 2 2 2 1 2 . . (Eq. 32) ( ) ( ) ( ) [ ]c a c b b a t w t y × + × + × × = . . 2 2 2 1 2 . (Eq. 33) ANL-EBS-MD-000003 REV 00 74 January 2000 [ ] ( ) ( ) ( ) [ ]2 2 2 2 2 2 c a c b b a c b t w a y × + × + × + × = . . . (Eq. 34) [ ] ( ) ( ) ( ) [ ]2 2 2 2 2 2 c a c b b a c a t w b y × + × + × + × = . . . (Eq. 35) [ ] ( ) ( ) ( ) [ ]2 2 2 2 2 2 c a c b b a b a t w c y × + × + × + × = . . . (Eq. 36) The maximum error in the corrosion rate is estimated by calculating numeric values of the partial derivatives from expected values of the independent variables, multiplication of each partial derivative by the corresponding error in independent variable ( .w, .., .t, .a, .b, and .c), and summation of the resulting products. The error based upon this method is shown in Table 16. Table 16. Summary of Error Analysis for Corrosion Rates Based Upon Weight Loss Measurements Assumed Weight Loss 0.0001 g 0.0010 g 0.0100 g .y .y .y Case Sample Configuration Exposure Time nm y-1 nm y-1 nm y-1 1 Crevice 6 month 12.25 12.95 19.86 2 Weight Loss 6 month 23.27 24.64 38.33 3 Crevice 12 month 6.00 6.29 9.17 4 Weight Loss 12 month 11.40 11.98 17.72 From the estimated errors given in Table 16 that are based on Tables 17 through 20, it is concluded that the typical uncertainty observed in weight loss and dimensional measurements prevent determination of corrosion rates less than approximately 38 nm y-1. The maximum uncertainty is estimated to be approximately 6 to 20 nm y-1 in the case of crevice samples and 11 to 38 nm in the case of weight loss samples. These estimates of probable error are believed to correspond to about one standard deviation (1 s). Therefore, any measured corrosion rate greater than 160 nm y-1 (4 s) should be easily distinguishable from measurement error. Any rate less than 160 nm y-1 guarantees that the WP outer barrier (wall thickness of 2 cm) will not fail by GC. ANL-EBS-MD-000003 REV 00 75 January 2000 Table 17. Error Analysis for LTCTF Corrosion Rates – Definitions Parameter Parameter Definition Units w Weight loss g . Density g cm-3 t Exposure time hr a Length in. b Width in. c Thickness in. a Length cm b Width cm c Thickness cm .y/.w Partial derivative or rate with respect to weight loss cm g-1 h-1 .y/.. Partial derivative of rate with respect to density cm4 g-1 h-1 .y/.t Partial derivative of rate with respect to exposure time cm h-2 .y/.a Partial derivative of rate with respect to length h-1 .y/db Partial derivative of rate with respect to width h-1 .y/.c Partial derivative of rate with respect to thickness h-1 .w Error in weight loss g .. Error in density g cm-3 .t Error in exposure time hr .a Error in length cm .b Error in width cm .c Error in thickness cm (.y/.w) × (.w) Weight loss product cm (.y/..) × (..) Density product cm (.y/.t) × (.t) Exposure time product cm (.y/.a) × (.a) Length product cm (.y/db) × (.b) Width product cm (.y/.c) × (.c) Thickness product cm .y Sum of all products cm h-1 .y Sum of all products µm y-1 .y Sum of all products nm y-1 ANL-EBS-MD-000003 REV 00 76 January 2000 Table 18. Error Analysis for LTCTF Corrosion Rates – Assume Weight Loss of 0.0001 Grams Parameter Crevice 6 month Weight Loss 6 month Crevice 12 month Weight Loss 12 month w 0.0001 0.0001 0.0001 0.0001 . 8.69 8.69 8.69 8.69 t 4296 4296 8760 8760 a 2.0000 2.0000 2.0000 2.0000 b 2.0000 1.0000 2.0000 1.0000 c 0.1200 0.1200 0.1200 0.1200 a 5.0800 5.0800 5.0800 5.0800 b 5.0800 2.5400 5.0800 2.5400 c 0.3048 0.3048 0.3048 0.3048 .y/.w 4.6338E-07 8.7964E-07 2.2725E-07 4.3139E-07 .y/.. 5.3324E-12 1.0122E-11 2.6151E-12 4.9642E-12 .y/.t 1.0786E-14 2.0476E-14 2.5942E-15 4.9245E-15 .y/.a 8.6331E-12 1.6435E-11 4.2337E-12 8.0601E-12 .y/db 8.6331E-12 3.1110E-11 4.2337E-12 1.5257E-11 .y/.c 1.6289E-11 4.4023E-11 7.9882E-12 2.1589E-11 .w 0.0003 0.0003 0.0003 0.0003 .. 0.1 0.1 0.1 0.1 .t 24 24 24 24 .a 0.00254 0.00254 0.00254 0.00254 .b 0.00254 0.00254 0.00254 0.00254 .c 0.00254 0.00254 0.00254 0.00254 (.y/.w) × (.w) 1.3902E-10 2.6389E-10 6.8174E-11 1.2942E-10 (.y/..) × (..) 5.3324E-13 1.0122E-12 2.6151E-13 4.9642E-13 (.y/.t) × (.t) 2.5887E-13 4.9142E-13 6.2260E-14 1.1819E-13 (.y/.a) × (.a) 2.1928E-14 4.1746E-14 1.0754E-14 2.0473E-14 (.y/db) × (.b) 2.1928E-14 7.9019E-14 1.0754E-14 3.8752E-14 (.y/.c) × (.c) 4.1374E-14 1.1182E-13 2.0290E-14 5.4837E-14 .y 1.3989E-10 2.6563E-10 6.8540E-11 1.3014E-10 .y 1.2255E-02 2.3269E-02 6.0041E-03 1.1401E-02 .y 1.2255E+01 2.3269E+01 6.0041E+00 1.1401E+01 ANL-EBS-MD-000003 REV 00 77 January 2000 Table 19. Error Analysis for LTCTF Corrosion Rates – Assume Weight Loss of 0.001 Grams Parameter Crevice 6 month Weight Loss 6 month Crevice 12 month Weight Loss 12 month w 0.0010 0.0010 0.0010 0.0010 . 8.69 8.69 8.69 8.69 t 4296 4296 8760 8760 a 2.0000 2.0000 2.0000 2.0000 b 2.0000 1.0000 2.0000 1.0000 c 0.1200 0.1200 0.1200 0.1200 a 5.0800 5.0800 5.0800 5.0800 b 5.0800 2.5400 5.0800 2.5400 c 0.3048 0.3048 0.3048 0.3048 .y/.w 4.6338E-07 8.7964E-07 2.2725E-07 4.3139E-07 .y/.. 5.3324E-11 1.0122E-10 2.6151E-11 4.9642E-11 .y/.t 1.0786E-13 2.0476E-13 2.5942E-14 4.9245E-14 .y/.a 8.6331E-11 1.6435E-10 4.2337E-11 8.0601E-11 .y/db 8.6331E-11 3.1110E-10 4.2337E-11 1.5257E-10 .y/.c 1.6289E-10 4.4023E-10 7.9882E-11 2.1589E-10 .w 0.0003 0.0003 0.0003 0.0003 .. 0.1 0.1 0.1 0.1 .t 24 24 24 24 .a 0.00254 0.00254 0.00254 0.00254 .b 0.00254 0.00254 0.00254 0.00254 .c 0.00254 0.00254 0.00254 0.00254 (.y/.w) × (.w) 1.3902E-10 2.6389E-10 6.8174E-11 1.2942E-10 (.y/..) × (..) 5.3324E-12 1.0122E-11 2.6151E-12 4.9642E-12 (.y/.t) × (.t) 2.5887E-12 4.9142E-12 6.2260E-13 1.1819E-12 (.y/.a) × (.a) 2.1928E-13 4.1746E-13 1.0754E-13 2.0473E-13 (.y/db) × (.b) 2.1928E-13 7.9019E-13 1.0754E-13 3.8752E-13 (.y/.c) × (.c) 4.1374E-13 1.1182E-12 2.0290E-13 5.4837E-13 .y 1.4779E-10 2.8126E-10 7.1830E-11 1.3670E-10 .y 1.2946E-02 2.4638E-02 6.2923E-03 1.1975E-02 .y 1.2946E+01 2.4638E+01 6.2923E+00 1.1975E+01 ANL-EBS-MD-000003 REV 00 78 January 2000 Table 20. Error Analysis for LTCTF Corrosion Rates – Assume Weight Loss of 0.01 Grams Parameter Crevice 6 month Weight Loss 6 month Crevice 12 month Weight Loss 12 month w 0.0100 0.0100 0.0100 0.0010 . 8.69 8.69 8.69 8.69 t 4296 4296 8760 8760 a 2.0000 2.0000 2.0000 2.0000 b 2.0000 1.0000 2.0000 1.0000 c 0.1200 0.1200 0.1200 0.1200 a 5.0800 5.0800 5.0800 5.0800 b 5.0800 2.5400 5.0800 2.5400 c 0.3048 0.3048 0.3048 0.3048 .y/.w 4.6338E-07 8.7964E-07 2.2725E-07 4.3139E-07 .y/.. 5.3324E-10 1.0122E-09 2.6151E-10 4.9642E-10 .y/.t 1.0786E-12 2.0476E-12 2.5942E-13 4.9245E-13 .y/.a 8.6331E-10 1.6435E-09 4.2337E-10 8.0601E-10 .y/db 8.6331E-10 3.1110E-09 4.2337E-10 1.5257E-09 .y/.c 1.6289E-09 4.4023E-09 7.9882E-10 2.1589E-09 .w 0.0003 0.0003 0.0003 0.0003 .. 0.1 0.1 0.1 0.1 .t 24 24 24 24 .a 0.00254 0.00254 0.00254 0.00254 .b 0.00254 0.00254 0.00254 0.00254 .c 0.00254 0.00254 0.00254 0.00254 (.y/.w) × (.w) 1.3902E-10 2.6389E-10 6.8174E-11 1.2942E-10 (.y/..) × (..) 5.3324E-11 1.0122E-10 2.6151E-11 4.9642E-11 (.y/.t) × (.t) 2.5887E-11 4.9142E-11 6.2260E-12 1.1819E-11 (.y/.a) × (.a) 2.1928E-12 4.1746E-12 1.0754E-12 2.0473E-12 (.y/db) × (.b) 2.1928E-12 7.9019E-12 1.0754E-12 3.8752E-12 (.y/.c) × (.c) 4.1374E-12 1.1182E-11 2.0290E-12 5.4837E-12 .y 2.2675E-10 4.3752E-10 1.0473E-10 2.0228E-10 .y 1.9863E-02 3.8327E-02 9.1744E-03 1.7720E-02 .y 1.9863E+01 3.8327E+01 9.1744E+00 1.7720E+01 ANL-EBS-MD-000003 REV 00 79 January 2000 6.5.4 Summary of General Corrosion Model Based upon these data and the associated error analysis presented in the following sections, a simple and defensible representation of the observed corrosion rates is proposed. This approach involves combining the distributions of rates calculated from weight loss and shown in Figures 23 and 25. These data are for “Weight Loss” and “Crevice” samples, respectively. It is assumed that no scale formation occurs. Therefore, all negative rates are eliminated, and the entire distribution can be assumed to be due to uncertainty. As shown in the resultant Figure 26, the rate at the 50th percentile is approximately 50 nm y-1; the rate at the 90th percentile is approximately 100 nm y-1; and the maximum rate is 731 nm y-1. About 10% of the values fall between 100 and 750 nm y-1. 0 10 20 30 40 50 60 70 80 90 100 0 100 200 300 400 500 600 700 800 Rate (nm/yr) Percentile (%) DTN: LL991208505924.099 Figure 26. GC Rates of Alloy 22 with Combined Data and Negative Values Neglected ANL-EBS-MD-000003 REV 00 80 January 2000 It would appear that the maximum value in the distribution of variability would be no greater than the maximum value in the distribution of uncertainty. Therefore, a conservative assumption would be to assume that the variability obeys a triangular distribution between zero and the maximum observed rate of 750 nm y-1. According to the literature (Evans et al. 1993) the distribution function is either b x a a c a b a x x F = = - - - = ) )( ( ) ( ) ( 2 (Eq. 37) or b x c c b a b x b x F = = - - - - = ) )( ( ) ( 1 ) ( 2 (Eq. 38) where c is the mode. The peak in the probability density function is about 2.0 and can be represented by either: b x a a c a b a x x f = = - - - = ) )( ( ) ( 2 ) ( (Eq. 39) or b x c c b a b x b x f = = - - - = ) )( ( ) ( 2 ) ( (Eq. 40) The mean and variance are given by: 3 ) ( c b a + + = µ (Eq. 41) and 18 2 2 2 bc ac ab c b a - - - + + = s (Eq. 42) If the probability density function is skewed to the lower values (as observed here), the following expression for c can be used: ) ( a b c - = a (Eq. 43) where the adjustable parameter alpha (a) is less than 0.5. ANL-EBS-MD-000003 REV 00 81 January 2000 6.5.5 Atomic Force Microscopy The AFM has been used to characterize the surface topographies of weight-loss coupons of Alloy 22 that had been exposed to various environments in the YMP’s LTCTF for one year. Having sub-nm vertical resolution, the AFM is an ideal tool for detecting extremely small penetrations in corrosion-resistant materials such as Alloy 22. As shown in Attachment I, Bedrossian and Fix have applied this technique to five Alloy 22 samples used for weight loss measurements (Bedrossian 1999). These samples include an unexposed control sample (DWA163), a sample exposed to aqueous phase SAW (DWA051), a sample exposed to vaporphase SAW (DWA048), a sample exposed to aqueous-phase SCW (DWA120), and a sample exposed to vapor-phase SCW (DWA117). The sample numbers are official designations of the YMP. After the samples were removed from the LTCTF, they were ultrasonically agitated in deionized water, acetone, and methanol for ten minutes each. The digital instruments DM3100 AFM was then used for imaging. Each set of data consists of a large-area scan (25 µm × 25 µm), followed by smaller-area details of the region displayed in the large-area scan. The gross surface topography is dominated by the machining grooves, with typical heights of several hundred nm and typical lateral periodicities of several µm features plainly visible on images of the control sample (DWA163, Figure 27). Samples removed from the LTCTF exhibit varying degrees of coverage by a deposit on top of this gross topography. The AFM images show that the most extensive deposit formation occurred on the sample exposed to aqueousphase SAW (DWA051, Figure 28). The next, most-extensive deposit formation occurred on the sample exposed to vapor-phase SAW (DWA048). X-ray Diffraction scans of all five coupons show that the deposit is predominantly a silicate or SiO2, with some NaCl appearing on the two samples which were in the SAW tank (Figure 29). Based upon both AFM and X-ray diffraction data, the two samples exposed to SCW showed lesser degrees of coverage by the silicate deposit. In some cases, depressions can be seen in the silicate deposit. However, it is not believed that any of these penetrate to the underlying metal. At the present time, there is insufficient data to quantitatively determine the extent of silicate removal from exposed Alloy 22 samples by acid cleaning. In the future, an effort will be made to collect sufficient quantitative information to quantitatively determine how much silicate remains on the surface after the acid cleaning procedure. In the mean time, a worst-case estimate of the impact of SiO2 on measured corrosion rates will be used. The formation of SiO2 deposits on the surface of the Alloy 22 could bias the distributions of GC rate shown in Sections 6.5.2 and 6.5.4. From various AFM images of Alloy 22 samples removed from the LTCTF, it appears that a typical deposit can have a thickness as great as 0.25 microns after 12 months of exposure. The resultant bias is then estimated. It is assumed that the deposit has the density of lechatelierite (amorphous SiO2), which is approximately 2.19 g cm-3 (Weast 1978, p. B-161). It is further assumed that the surface is completely and uniformly covered by this deposit. The estimated surface areas of the weight-loss and crevice samples are 30.65 and 57.08 cm2, respectively (4.75 and 8.85 in2, respectively). Consequently, the deposit thickness translates into a mass change of 1.678 and 3.125 mg for weight-loss and crevice samples, respectively, after 12 months of exposure. Equation 24 is then applied to determine the impact of such a positive mass change on the calculated GC rate. In the case of the weight loss sample, the estimated bias is 0.063 microns per year (63 nm y-1): ANL-EBS-MD-000003 REV 00 82 January 2000 k = 104 µmcm-1 .w = 1.678 ×10-3 gm area = 30.645 cm 2 t = 1.0 y .= 8.69 gm cm-3 . dp dt . . . . = 104 µmcm-1 ( )1.678 ×10-3 gm ( ) 30.645 cm 2 ( )1.0 yr ( )8.69 gm cm-3 ( )= 0.063 µm y-1 In the case of the crevice sample, the result is the same: k = 104 µmcm-1 .w = 3.125 × 10-3 gm area = 57.078 cm 2 t = 1.0 y .= 8.69 gm cm-3 . dp dt . . . . = 104 µm cm-1 ( )3.125 ×10-3 gm ( ) 57.078 cm 2 ( )1.0 yr ( )8.69 gm cm-3 ( )= 0.063 µm y-1 The distributions of GC rate shown in Sections 6.5.2 and 6.5.4 can be corrected for the maximum bias due to SiO2 deposit formation by adding a constant value of 63 nm y-1 to each estimated value of the GC rate. This is equivalent to shifting the curves shown in Figures 23, 25, and 26 to the right by 63 nm y-1. ANL-EBS-MD-000003 REV 00 83 January 2000 NOTE: Bedrossian (1999) Figure 27. AFM Image of Alloy 22 Control Sample NOTE: Bedrossian (1999) Figure 28. AFM Image of Alloy 22 Sample Removed from LTCTF ANL-EBS-MD-000003 REV 00 84 January 2000 800 600 400 200 0 36 34 32 30 28 26 24 22 20 2-Theta (deg) SiO2 SiO2 NaCl DWA163 (spectrum z03309) DWA051 (spectrum z03320) DWA048 (spectrum z03316) NOTE: Bedrossian (1999) Figure 29. X-ray Diffraction Pattern of Silicate Deposit on Sample Exposed to LTCTF The AFM has been used to examine areas inside and outside of Alloy 22 crevices exposed for 12 months to SCW at 90°C. Though the images were obtained with a welded sample (DCB100), the unwelded area was imaged with the AFM. Figure 30 shows two optical micrographs of the sample surface near the hole (0.312 inch diameter). The bottom image is a 10× magnification of the top image. There is some discoloration underneath the crevice, but no evidence of penetration. Figure 31 shows AFM data for an area 1.5 mm outside the crevice, plotted in the form of a probability density function for vertical distance. This distribution peaks at 597 nm. The corresponding three-dimensional image is shown in Figure 32. For comparison, Figure 33 shows AFM data for an area 1.5 mm inside the crevice, plotted in the form of a probability density function for vertical distance. This distribution peaks at 549 nm, very close to that determined for the area outside of the crevice. The corresponding three-dimensional image is shown in Figure 34. AFM line scans perpendicular to the edge, along the outside area, along the inside area, and along an area on an unexposed control sample are compared in Figure 35. There appears to be no significant difference between the roughness of the four areas that were examined. Because it has been observed that corrosion tends to roughen the surface, it is concluded that there is no more attack inside the crevice than outside. ANL-EBS-MD-000003 REV 00 85 January 2000 NOTE: The top image has a relative magnification factor of 1X. The bottom image has a relative magnification factor of 10X. Figure 30. Photographs of Alloy 22 Crevice Area after 12-month Exposure to SCW Aqueous Phase at 90°C (DCA101). diameter of hole = 0.312 inches ANL-EBS-MD-000003 REV 00 86 January 2000 0 500 1000 1500 2000 2500 3000 3500 0 100 200 300 400 500 600 700 800 900 1000 Vertical Distance (nm) Frequency (pixels) Distribution of Vertical Distance 1.5 mm Outside of Crevice (DCB100_4) 597 nm NOTE: Bedrossian (1999) Figure 31. Histogram of AFM Measurements of Vertical Distance made 1.5 mm outside of Crevice (DCB100_4) NOTE: Bedrossian (1999) Figure 32. Three-dimensional AFM Image taken 1.5 mm outside of Crevice (DCB100_4) ANL-EBS-MD-000003 REV 00 87 January 2000 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 100 200 300 400 500 600 700 800 900 1000 Vertical Distance (nm) Frequency (pixels) Distribution of Vertical Distance 1.5 mm Inside of Crevice (DCB100_5) 549 nm NOTE: Bedrossian (1999) Figure 33. Histogram of AFM Measurements of Vertical Distance made 1.5 mm inside of Crevice (DCB100_5) NOTE: Bedrossian (1999) Figure 34. Three-dimensional AFM Image taken 1.5 mm inside of Crevice (DCB100_5) ANL-EBS-MD-000003 REV 00 88 January 2000 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 0 10 20 30 40 50 60 70 80 90 100 Horizontal Distance (micrometers) Adjusted Vertical Distance (nanometers) (DCB100_1) across crevice edge (DCB100_4) 1.5 mm outside crevice (DCB100_5) 1.5 mm inside crevice (PB99067.019) control sample AFM Line Scan Across Surfaces on Crevice Sample 1 10 100 1000 10000 0 10 20 30 40 50 60 70 80 90 100 Horizontal Distance (micrometers) Adjusted Vertical Distance (nanometers) (DCB100_1) across crevice edge (DCB100_4) 1.5 mm outside crevice (DCB100_5) 1.5 mm inside crevice (PB99067.019) control sample AFM Line Scan Across Surfaces on Crevice Sample NOTE: Bedrossian (1999) The top graph has a linear scale. The bottom graph has a logarithmic scale. Figure 35. A Comparison of AFM Line Scans across Different Regions of Exposed Crevice Sample, with Comparison to Line Scan on Surface of Unexposed Control Sample ANL-EBS-MD-000003 REV 00 89 January 2000 A study of four test coupons of Alloy 22 removed from the LTCTF after one year showed varying degrees of coverage by silicate deposits but no evidence of localized corrosion by pitting. The distributions of GC rate shown in Sections 6.5.2 and 6.5.4 can be corrected for the maximum bias due to SiO2 deposit formation by adding a constant value of 63 nm y-1 to each estimated value of the GC rate. This is equivalent to shifting the curves shown in Figures 23, 25, and 26 to the right by 63 nm y-1. The AFM has been used to examine areas inside and outside of Alloy 22 crevices exposed to SCW at 90°C for 12 months. AFM line scans perpendicular to the edge, along the outside area, along the inside area, and along an area on an unexposed control sample are compared in this AMR. There appears to be no significant difference between the roughness of the four areas that were examined. Since it has been observed that corrosion tends to roughen the surface, it is concluded that there is no more attack inside the crevice than outside. 6.5.6 Dissolved Oxygen in the Long Term Corrosion Test Facility Corrosion rates in the LTCTF may depend upon the concentration of dissolved oxygen because the cathodic reduction of oxygen may be required to depolarize anodic dissolution reactions. The anodic dissolution of a metal requires a corresponding amount of cathodic reduction. Typically, dissolved oxygen or hydrogen ion is reduced. However, as previously discussed, other reactants such as hydrogen peroxide (due to gamma radiolysis) can also be reduced. Figure 36 shows a comparison of dissolved oxygen measurements in LTCTF to published data for synthetic geothermal brine (Cramer 1974). The published data spans the range of temperature from 20 to 300°C, and spans the range of oxygen partial pressures from 1 to 30 psi. Note that the partial pressure of oxygen in the atmosphere is about 3 psi. The points representing measurements from the LTCTF tanks are superimposed upon the published data. Clearly, the SDW, SCW, and SAW appear to be saturated (4-10 ppm dissolved oxygen). 0.1 1 10 100 0 50 100 150 200 250 300 350 Temperature (Centigrade) Dissolved Oxygen (ppm) 1 psi 3 psi 10 psi 30 psi SDW Pair # 1 (3 psi) SDW Pair # 2 (3 psi) SDW Pair # 3 (3 psi) SDW Pair # 4 (3 psi) SCW Pair # 1 (3 psi) SCW Pair # 2 (3 psi) SCW Pair # 3 (3 psi) SCW Pair # 4 (3 psi) SAW Pair # 1 (3 psi) SAW Pair # 2 (3 psi) Air DTN: LL990610605924.079 NOTE: Cramer (1974) Figure 36. Comparison of Dissolved Oxygen Measurements in LTCTF to Data for Synthetic Geothermal Brine ANL-EBS-MD-000003 REV 00 90 January 2000 6.6 CREVICE CORROSION 6.6.1 Scenarios Leading to Crevice Formation At points of contact between the WP and other solid objects, crevices form occluded geometries, which lead to differential aeration of the crevice solution (electrolyte). Dissolved oxygen can become depleted deep within the crevice, while the oxygen concentration near the crevice mouth remains relatively high. Cathodic reduction of dissolved oxygen at the crevice mouth may create a sufficiently high electrochemical potential to drive anodic processes inside the crevice, thereby causing an anodic current to flow along the crevice towards the crevice mouth. Under realistic repository conditions, it is believed that the walls of the Alloy 22 crevice will remain passive. The potential at the mouth of a crevice is expected to be well below the threshold for localized attack, as determined with CP measurements. Anodic processes inside the crevice are, therefore, expected to occur at a rate that corresponds to the local passive current density. Two primary electrochemical processes can lead to acidification of the solution in a passive crevice, (1) the preferential transport of anions into the crevice from the mouth, driven by the electric field that accompanies the crevice current and (2) hydrolysis reactions of dissolved metal cations. Based upon experimental work with passive crevices without buffer, it is believed that the applied potentials required for significant acidification (pH<5) are not plausible. A minimum crevice pH of approximately 5 is assumed. Additional experimental work of the type discussed here is required to further substantiate this preliminary conclusion. 6.6.2 Crevice Chemistry and Lowering of Local pH The hydrolysis of dissolved metal in crevices can lead to the accumulation of H+ and the corresponding suppression of pH. For example, pH < 2 has been observed in crevices made of stainless steel, as discussed by Sedriks (1996). Metal ions produced by anodic dissolution are assumed to undergo the following hydrolysis reactions, as discussed by Oldfield and Sutton (1978): Fe3+ + H2O . . . Fe(OH)+ + H+ (Eq. 44) + + + + .. . + H OH Fe O H Fe 2 2 3 ) ( (Eq. 45) + + + + .. . + H OH Ni O H Ni ) ( 2 2 (Eq. 46) + + + + .. . + H OH Cr O H Cr 2 2 3 ) ( (Eq. 47) + + + + .. . + H OH Cr O H OH Cr 2 ) ( ) ( 2 2 (Eq. 48) If the dissolved metals exceed the solubility limits, precipitation will occur: - + + .. . OH Fe s OH Fe 2 ) ( ) ( 2 2 (Eq. 49) - + + .. . OH Ni s OH Ni 2 ) ( ) ( 2 2 (Eq. 50) ANL-EBS-MD-000003 REV 00 91 January 2000 - + + .. . OH Cr s OH Cr 3 ) ( ) ( 3 3 (Eq. 51) Precipitation of hydroxides are favored at more alkaline pH levels. In the case of Alloy 22, the hydrolysis of other dissolved metals such as molybdenum and tungsten ions may be important. The Oldfield-Sutton model does not account for the role of HCl in the crevice on destabilization of the passive film. 6.6.3 Chloride Transport by Electromigration Chloride anion will be driven into the crevice by the potential gradient, as discussed in the literature (Pickering and Frankenthal 1972; Galvele 1976). The corresponding concentration in the crevice is: Cl- [ ]= Cl - [ ]0 exp - F RT F x ( ) . . . . (Eq. 52) where [Cl-]0 is the concentration at the crevice mouth, F(x) is the potential in the crevice relative to that at the mouth, and (x) is the distance from the crevice mouth. Field-driven electromigration of Cl- (and other anions) into crevice must occur to balance cationic charge associated with H+ ions, as well as the charge associated with Fe2+, Ni2+, Cr3+, and other cations. If such conditions do develop inside Alloy 22 crevices, the stage might be set for an accelerated attack of this material by localized corrosion or SCC. 6.6.4 Deterministic Models of the Crevice A detailed deterministic model has been developed to calculate the spatial distributions of electrochemical potential and current density in WP crevices, as well as transient concentration profiles of dissolved metals and ions (Farmer and McCright 1998; Farmer et al. 1998). These quantities are calculated with the transport equations, which govern electromigration, diffusion, and convective transport. In cases with strong supporting electrolyte, electromigration can be ignored (Newman 1991). First, the axial current density along the length of the crevice is calculated by integrating the wall current density. The electrode potential along the length of the crevice can then be calculated from the axial current density. This technique is similar to that employed in other models (Nystrom et al. 1994). Such models show that the electrochemical potential decreases with increasing distance into the crevice. Therefore, the potential should never be more severe (closer to the threshold for LC) than at the crevice mouth. The partial differential equations that define transient concentrations in the crevice require determination of the potential gradient, as well as the local generation rates for dissolved species. The concentrations of dissolved metals at the crevice mouth are assumed to be zero. Computations are facilitated by assuming that the crevices are symmetric about a mirror plane where the flux is zero. This model has been used to estimate the extent of pH suppression in WP crevices due to the simultaneous hydrolysis and transport of dissolved Fe, Ni, Cr, Mo, and W. 6.6.5 Experimental Determinations of Crevice pH and Current The local crevice environments for Alloy 22 and other relevant materials are being determined experimentally. This procedure is described in AP-E-20-81, Revision 1. Crevices have been ANL-EBS-MD-000003 REV 00 92 January 2000 constructed from square metallic samples, 2 inches on each side and 1/8 inch thick (same size as crevice samples used in the LTCTF). The samples are masked with plastic tape, thereby forming an exposed square area, 1.7 inches on each side. The exposed area is placed underneath a clear plastic window with an access port for a pH sensor in the center. In this case, the sensor is a miniature reference electrode separated from the crevice solution with a thin glass membrane. A second pH sensor is located at the mouth of the crevice, in close proximity to a saturated calomel reference electrode (SCE). The use of in situ sensors to determine crevice pH has also been described by Sridhar and Dunn (1994). In parallel experiments by Farmer et al. (1998), paper strips with a pH-sensitive dye (pH paper) have been sandwiched between the clear plastic window and photographed with a digital electronic camera in a time-lapse mode to add confidence to the measurements made with pH sensors. Spectroscopic-grade graphite counter electrodes are also placed in the electrolyte lying outside the mouth of the crevice. A potentiostat is then used to control the electrochemical potential at the mouth of the crevice. Temperature, potential, current, and pH is then recorded electronically during the course of the experiment. Measurements of pH inside a crevice formed from 316L stainless steel are shown in Figure 37. The electrolyte was 4M NaCl and was maintained at ambient temperature. Since this electrolyte contains no buffer ions, it is considered to be a far more severe medium than those representative of various concentrations of J-13 well water. The electrochemical potential at the mouth was maintained at 200 mV versus Ag/AgCl. Crevice corrosion could be seen initiating near the crevice mouth and propagating towards the pH sensor, which was located about 0.5 cm inside the crevice mouth. When the corrosion front reached the pH sensor, the pH dropped from the initial value (pH~7) to a very low value (pH~1). The fixed one-liter volume of electrolyte outside of the crevice became slightly alkaline. The pH of this solution reached a maximum (pH~10) and then fell to a slightly lower steady-state value (pH~9). Active corrosion inside the crevice is evident since the color of the crevice solution becomes emerald green. In similar experiments with 316L exposed to SCW, no significant lowering of the pH was observed. Measurements of pH inside crevices formed with Alloy 22 surfaces are shown in Figures 38 through 42. Figure 39 shows the evolution of pH in a crevice with a potential of 800 mV versus Ag/AgCl applied at the mouth. The electrolyte was 4M NaCl and was maintained at ambient temperature. The Alloy 22 surface remained passive underneath the window, with no visible signs of localized attack. However, the passive current flow from within the crevice was sufficient to cause the pH to be immediately lowered from the initial value (pH~6.5) to a minimum value (pH~3.3), after which the pH gradually increased over several hours (pH~4.5). The fixed one-liter volume of electrolyte outside of the crevice became slightly alkaline (pH~8.3) before the data acquisition was started and dropped gradually over several hours (pH~7). The lowering of pH inside of passive Alloy 22 crevices with high-applied potential has been verified by independent technique-development tests with indicator paper, as discussed in AP-E-20-81 Rev. 1. Figures 38 through 41 illustrate the effect of increasing the applied potential above the threshold required for localized breakdown of the passive film. As shown in Figure 39, an applied potential of 1100 mV can drive the pH to extremely low levels (pH~0.2) in Alloy 22 crevices. Figures 40 and 41 show the effect of incremental changes in applied potential on both crevice pH and crevice current. At an applied potential of 400 mV, the steady-state crevice pH remained close to neutrality (pH~6.1). As the potential was stepped to 1000 mV, which is slightly above the repassivation potential measured by Gruss et al. (1998), the crevice current ANL-EBS-MD-000003 REV 00 93 January 2000 increased dramatically and the pH dropped below one. At an applied potential of 1100 mV, extreme localized attack of the Alloy 22 was observed at the crevice mouth, with a crevice pH slightly less than zero. At the end of the experiment, the crevice sensor was immediately submersed in a buffer solution (pH 7) and shown to be in good calibration (virtually no drift during test). Figure 42 shows the effect of buffer ions on crevice chemistry. In this case, SCW was used as the electrolyte. Even at an applied potential of 800 mV, no significant lowering of the pH was observed. The Alloy 22 inside the crevice appeared to be unchanged from its initial state, with no evidence of localized attack. Figure 43 is a summary of several experiments where crevice pH was determined in situ as a function of applied potential. These data are represented by the following polynomial: 2 2 1 0 x b x b b y + + = (Eq. 53) where x is the potential applied at the crevice mouth (mV versus Ag/AgCl) and y is the steadystate pH inside the crevice. Coefficients for the above equation are summarized in Table 21, representing both Alloy 22 and 316L in under a broad range of conditions. The correlations for 4M NaCl and SCW should be used to bound the crevice pH, using linear interpolation between the two limits, based upon the concentration of buffer ion. Table 21. Coefficients for the Correlation of Crevice pH with Applied Potential Material Medium Crevice Spacer b0 b1 b2 R2 (µm) Alloy 22 4M NaCl 110 7.2716 -0.0012 -5.0E-06 0.9782 Alloy 22 4M NaCl 540 7.0227 -0.0015 -4.0E-06 ~1 Alloy 22 SCW 540 8.276 0.0003 0.9646 316L 4M NaCl 540 1.035 -0.00001 0.0005 316L SCW 540 8.1175 -0.00006 ~1 DTN: LL991208605924.100 In summary, there was no visible evidence of localized corrosion of the metal inside the crevice at applied potentials less than the threshold. However, even though the crevice remained passive, the passive current density and imposed electric field within the crevice was sufficient to cause significant acidification. In many of the experiments described here, both the applied potential and the test medium are more severe than those expected in the repository. However, the temperature of aqueous solutions on the WP surface may be significantly higher (120°C). Work is in progress to obtain comparable data at higher temperature. The experimental data support published numerical simulations (Farmer et al. 1998; Farmer et al. 1999). ANL-EBS-MD-000003 REV 00 94 January 2000 0 1 2 3 4 5 6 7 8 9 10 0 120 240 360 480 600 720 840 960 1,080 1,200 Time (minutes) Crevice pH -400 -200 0 200 400 600 800 1,000 1,200 Potential at Mouth (mV vs. Ag/AgCl) Inside of Crevice (pH) Mouth of Crevice (pH) Potential at Mouth (mV vs. Ag/AgCl) DTN: LL990610505924.078 Figure 37. Stainless Steel 316L, 4M NaCl, 200 mV and 23 °C, Crevice pH Versus Time 0 1 2 3 4 5 6 7 8 9 10 0 120 240 360 480 600 720 840 960 1,080 1,200 1,320 1,440 Time (minutes) Crevice pH -600 -400 -200 0 200 400 600 800 1,000 1,200 Potential at Mouth (mV vs. Ag/AgCl) Inside of Crevice (pH) Mouth of Crevice (pH) Mouth Potential (mV vs. Ag/AgCl) DTN: LL990610505924.078 Figure 38. Alloy 22, 4M NaCl, 800 mV and 23 °C, Crevice pH Versus Time ANL-EBS-MD-000003 REV 00 95 January 2000 0 1 2 3 4 5 6 7 8 9 10 0 240 480 720 960 1,200 1,440 Time (minutes) Crevice pH -400 -200 0 200 400 600 800 1,000 1,200 Potential at Mouth (mV vs. Ag/AgCl) Inside of Crevice (pH) Mouth of Crevice (pH) Potential at Mouth (mV vs. Ag/AgCl) DTN: LL990610505924.078 Figure 39. Alloy 22, 4M NaCl, 1100 mV and 20 °C, Crevice pH Versus Time -2 -1 0 1 2 3 4 5 6 7 8 9 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 Time (minutes) Crevice pH -400 -200 0 200 400 600 800 1,000 1,200 Potential at Mouth (mV vs. Ag/AgCl) Inside of Crevice (pH) Mouth of Crevice (pH) Potential at Mouth (mV vs. Ag/AgCl) Sensors return to original values at end of experiment DTN: LL990610505924.078 Figure 40. Alloy 22, 4M NaCl at 23 °C, Crevice pH Versus Time ANL-EBS-MD-000003 REV 00 96 January 2000 1 10 100 1,000 10,000 100,000 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 Time (minutes) Crevice Current (microamps) -200 0 200 400 600 800 1,000 1,200 Potential at Mouth (mV vs. Ag/AgCl) Current at Mouth (microamps) Potential at Mouth (mV vs. Ag/AgCl) Sample returns to corrosion potential DTN: LL990610505924.078 Figure 41. Alloy 22, 4M NaCl at 23 °C, Crevice Current Versus Time 0 1 2 3 4 5 6 7 8 9 10 0 600 1,200 1,800 2,400 3,000 Time (minutes) Crevice pH -400 -200 0 200 400 600 800 1,000 1,200 Potential at Mouth (mV vs. Ag/AgCl) Inside of Crevice (pH) Mouth of Crevice (pH) Potential at Mouth (mV vs. Ag/AgCl) DTN: LL990610505924.078 Figure 42. Alloy 22, SCW at 23 °C, Crevice pH Versus Time ANL-EBS-MD-000003 REV 00 97 January 2000 pH = - 5x10-6 E2 - 0.0012 E + 7.2716 R2 = 0.9782 pH = 6x10-5 E + 8.1175 R2 = 1 pH = 3.0586e-0.0012 E R2 = 0.9608 pH = 0.0003 E + 8.276 R2 = 0.9646 pH = - 4x10-6 E2 - 0.0015 E + 7.0227 R2 = 1 pH = - 1x10-5 E + 1.035 R2 = 0.0005 -1 0 1 2 3 4 5 6 7 8 9 10 0 200 400 600 800 1000 1200 Potential at Crevice Mouth (mV vs. Ag/AgCl) Crevice pH 316L 0.54 mm Satd. KCl 20 C 316L 0.54 mm 4M NaCl 20 C 316L 0.54 mm SCW 20 C 316L 0.54 mm 4M NaCl & SCW 20 C C22 0.54 mm 4M NaCl 20 C C22 0.54 mm SCW 20 C C22 0.54 mm 4M NaCl & SCW 20 C C22 0.11 mm 4M NaCl 20 C C22 0.11 mm 4M NaCl 20 C Poly. (C22 0.11 mm 4M NaCl 20 C) Poly. (316L 0.54 mm SCW 20 C) Expon. (316L 0.54 mm Satd. KCl 20 C) Linear (C22 0.54 mm SCW 20 C) Poly. (C22 0.54 mm 4M NaCl 20 C) Linear (316L 0.54 mm 4M NaCl 20 C) 316L & C-22 with buffer C-22 with high Cl- & no buffer 316L with high Cl- & no buffer Buffer ions precipitate at elevated temperature DTN: LL990610305924.076 Figure 43. Determination of Crevice pH for WP Materials 6.6.6 Estimated Rate of Localized Corrosion If the threshold potential for localized attack is exceeded, a corrosion rate representative of LC must be assumed. Due to the outstanding corrosion resistance of Alloy 22, very little data exists for such localized corrosion under plausible conditions. Work originally published by Asphahani (1980) and later reviewed by Gdowski (1991) indicates that the corrosion rate of Alloy 22 in 10 wt% FeCl3 at 75°C might be as high as 12.7 µm per year. This rate is significantly higher than those measured in the LTCTF and may be representative of the types of rates expected for LC, including crevice corrosion. In a solution composed of 7 vol% H2SO4, 3 vol% HCl, 1 wt% FeCl3, and 1 wt% CuCl2, a penetration rate of 610 µm per year was observed at 102°C. From 9.12 (Sedriks 1996), the corrosion rate of Alloy C-276 in dilute HCl at the boiling point is somewhere between 5 and 50 mils per year (127 and 1270 µm per year). Comparable rates would be expected for Alloy 22. The highest passive current density found in Figures 15 through 18 is approximately 10 µA cm-2, which corresponds to a corrosion rate of approximately 100 µm per year. For the time being, it is expected that the logarithm of the localized corrosion rate of Alloy 22 is normally distributed, as shown in Table 22. This distribution reasonably bounds those extreme penetration rates found in the literature and is centered around the rate corresponding to the passive current density. Table 22. Distribution of LC Rates for Alloy 22 Percentile (%) Localized Corrosion Rate (µm per year) 0th 12.7 50th 127 100th 1270 NOTE: Asphahani (1980); Gdowski (1991); Sedriks (1996) ANL-EBS-MD-000003 REV 00 98 January 2000 6.7 EFFECT OF AGING AND PHASE INSTABILITY ON CORROSION The WP surface temperature is always below 300°C. With this constraint, the impact of aging and phase instability on the corrosion of Alloy 22 will be insignificant. An extrapolation of the curves given in the companion AMR on aging and phase stability does not indicate that the phase stability of Alloy 22 base metal will be a problem at less than about 300°C (CRWMS M&O 2000b). However, it must be emphasized that such estimates are preliminary and uncertain. Much additional work is needed in this area. Rebak et al. have investigated the effects of hightemperature aging on the corrosion resistance of Alloy 22 in concentrated hydrochloric acid. However, due to the temperature used to age the samples (922-1033 K) and the extreme test media used (boiling 2.5% HCl and 1 M HCl at 339 K), these data are not considered relevant to performance assessment for the repository. This data will soon be published by R. B. Rebak, N. E. Koon, and P. Crook in an article entitled “Effect of High Temperature Aging on the Electrochemical Behavior of C-22 Alloy.” This paper will appear in the Proceedings of the 50th Meeting of the International Society of Electrochemistry, which documents a conference held in Pavia, Italy, in September 1999. 6.7.1 Corrosion Testing of Aged Samples in Standard Simulated Acidic Concentrated Water and Simulated Concentrated Water Test Media Samples of Alloy 22 were aged at 700°C for either 10 or 173 hours. The corrosion resistance of these aged samples is compared to that of base metal in various standardized test media. Figure 44 shows a comparison of CP curves for base metal and thermally aged material in SAW at 90°C. Both curves exhibit generic type 1 behavior. In this case, aging appears to shift the corrosion potential to less noble values from -176 to -239 mV verses a standard Ag/AgCl reference electrode. The passive current density may be increased slightly, which would be indicative of a slight increase in corrosion rate. The highest non-equilibrium passive current observed for the base metal is approximately 4 µA cm-2 compared to approximately 10 µA cm-2 for fully aged material. The effect of thermal aging on the corrosion rate is accounted for in the enhancement factor, Gaged, and is based upon a ratio of the non-equilibrium current densities for base metal and aged material. effective aged effective dt dp G dt dp × = (Eq. 54) The value of Gaged for base metal is approximately one (Gaged ~ 1), whereas the value of Gaged for fully aged material is larger (Gaged ~ 2.5). Material with less precipitation than the fully aged material would have an intermediate value of Gaged (1 = Gaged = 2.5). Figure 45 shows a comparison of CP curves for base metal and thermally aged material in SCW at 90°C. In this case, aging also appears to shift the corrosion potential to less noble values from -237 to somewhere between -328 and -346 mV verses a standard Ag/AgCl reference electrode. In all three cases, the anodic oxidation peak that is characteristic of generic type 2 behavior is observed. ANL-EBS-MD-000003 REV 00 99 January 2000 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00 -400 -200 0 200 400 600 800 1000 1200 1400 Potential (mV vs. Ag/AgCl) Current Density (A/cm 2) DEA002 base metal DEA201 aged for 173 hours at 700 C Ecorr = -239 mV Ecorr = -176 mV Effect of Aging on Corrosion of Alloy 22 in SAW at 90 C Scan Direction DTN: LL991208705924.101 Figure 44. Effect of Thermal Aging for 173 Hours at 700°C on the Corrosion Resistance of Alloy 22 in SAW at 90°C (DEA002 and DEA201) 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00 -600 -400 -200 0 200 400 600 800 1000 1200 1400 Potential (mV vs. Ag/AgCl) Current Density (A/cm 2) DEA016 base metal DEA202 aged 173 hours at 700 C DEA202 aged 173 hours at 700 C Ecorr = -237 mV Ecorr = -346 mV Ecorr = -328 mV Effect of Aging on Corrosion of Alloy 22 in SCW at 90 C Scan Direction DTN: LL991208705924.101 Figure 45. Effect of Thermal Aging for 173 hours at 700°C on the Corrosion Resistance of Alloy 22 in SCW at 90°C (DEA016, DEA202 and DEA203) ANL-EBS-MD-000003 REV 00 100 January 2000 6.7.2 Worst-case Test for Aged Samples CP curves for base metal and thermally aged material in a new test medium of interest, BSW-13 at 110°C, are also compared. These data represent a worst-case test for Alloy 22, a combination of extreme thermal aging, extreme water chemistry, and a temperature approaching the boiling point. The BSW composition was established on the basis of results from a distillation experiment (CRWMS M&O 2000a). The total concentration of dissolved salts in the starting liquid was approximately five times more concentrated than that in the standard SCW solution. It was prepared by using five times the amount of each chemical that is specified for the preparation of SCW. After evaporation of ~90% of the water from the starting solution, the residual solutions reaches the highest chloride concentration and has a boiling point of ~111°C. The resultant BSW solution contains (sampled at 111°C) 9% chloride, 9% nitrate, 0.6% sulfate, 0.1% fluoride, 0.1% metasilicate, 1% TIC (total inorganic carbon from carbonate and bicarbonate), 5% potassium ion, and 11% sodium ion. A recipe for preparing synthetic BSW is shown below in Table 23. Table 23. Initial BSW Solution Recipe Chemical Quantity (g) Na2CO3 (anhydrous) 10.6 KCl 9.7 NaCl 8.8 NaF 0.2 NaNO3 13.6 Na2SO4 (anhydrous) 1.4 H2O 55.7 pH 11.3 (measured at room temperature) DTN: LL991213805924.110 The synthetic BSW solution represented by Table 23 has been slightly modified for these and other corrosion tests, yielding BSW-11, BSW-12, and BSW-13. The three solutions have pH values of approximately 13, 12, and 11 respectively. All BSW-type solutions contain 9% chloride, 9% nitrate, 0.6% sulfate, and 0.1% fluoride. Sodium and potassium ions are used to balance the charge. More specifically, each testing solution contains 8.7 g KCl, 7.9 g NaCl, 0.2 g NaF, 13.6 g NaNO3, and 1.4 g Na2SO4 (anhydrous). The pH 13 solution (BSW-13) was prepared by adding 65 mL of water and 2.0 mL of the 10 N NaOH to the chemicals (total weight = 100 g). The measured pH was 13.13. The pH 12 solution (BSW-12) was prepared by adding 66 mL of water and 2.0 mL of the 1 N NaOH to the chemicals. The measured pH was 12.25. The pH 11 solution (BSW-11) was prepared by adding 66 mL of water and 2.0 mL of the 0.1 N NaOH to the chemicals. The measured pH was 11.11. These recipes are summarized below in Table 24. It should be pointed that the modified BSW solutions are not buffered. ANL-EBS-MD-000003 REV 00 101 January 2000 Table 24. Modified BSW Solution Recipes BSW-13 BSW-12 BSW-11 Chemical Quantity Quantity (g) Quantity (g) KCl 8.7 g 8.7 g 8.7 g NaCl 7.9 g 7.9 g 7.9 g NaF 0.2 g 0.2 g 0.2 g NaNO3 13.0 g 13.0 g 13.0 g Na2SO4 (anhydrous) 1.4 g 1.4 g 1.4 g H2O (deionized) 66 ml 66 ml 66 ml 10N NaOH 2 ml 1N NaOH 2 ml 0.1N NaOH 2 ml CO2 partial pressure 0 0 0 pH (measured at room temperature) 13.13 12.25 11. 11 DTN: LL991213805924.110 NOTE: The CO2 partial pressure can be minimized by either scrubbing laboratory air or purchasing CO2 free air. In order to add some soluble silica to the solution, the BSW solution recipe was later revised to contain 4.0 g (~1% metasilicate) by adding sodium metasilicate (Na2SiO3 .9H2O). With the addition of the metasilicate, the pH was increased from 11.3 to 13 as measured at room temperature. It has been noted that the pH of aqueous solutions is dependent on the partial pressure of gaseous CO2. The implication of this is that unless many constraints are taken to control the pH of the BSW solution, the pH may vary with test conditions. It is not known with what partial pressure of CO2 that the revised BSW solution is in equilibrium. In order to conduct a long time testing (few months to a year), the testing environments should be stable. It was decided that to make stable testing solutions, carbonate and silicates should not be added to the test solution as both species can affect solution pH. Instead, sodium hydroxide will be used to maintain the higher pH values of solution. Gaseous CO2 must be also removed from the air passing above the solution because, as noted above, it will affect the solution pH. With no gaseous CO2 in contact with the solution and no carbonate/bicarbonate and silicates in solution, the testing environments should be stable. In tests with BSW-13 (Figure 46), aging also appears to shift the corrosion potential to less noble values. A sample aged for only 10 hours has a corrosion potential of only -227 mV verses a standard Ag/AgCl reference electrode, whereas a sample aged for 173 hours has a corrosion potential of -372 mV relative to the same reference. The difference Ecritical-Ecorr is about 800 mV for an aged sample in either SAW and BSW. The non-equilibrium current densities (corrosion rates) at 0 mV are also similar. However, more quantitative test are required for any definitive statements regarding corrosion rate. ANL-EBS-MD-000003 REV 00 102 January 2000 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00 -800 -600 -400 -200 0 200 400 600 800 1000 1200 1400 Potential (mV vs. Ag/AgCl) Current Density (A/cm2) DEA159 aged for 10 hours at 700 C DEA209 aged for 173 hours at 700 C Ecorr = -372 mV Ecorr = -227 mV Effect of Aging on Corrosion of Alloy 22 in BSW at 110°C Scan Direction DTN: LL991208705924.101 Figure 46. Effect of Thermal Aging at 700°C on the Corrosion Resistance of Alloy 22 in BSW-13 at 110°C (DEA159 and DEA209) 6.7.3 Accounting for Overall Effect of Thermal Aging on Corrosion A fully aged sample of Alloy 22 appears to exhibit a less noble corrosion potential, shifted in the cathodic direction by approximately: 63 mV in the case of SAW at 90°C; 109 mV in the case of SCW at 90°C; and by more than 100 mV in the case of BSW at 110°C. It is assumed that Ecorr can be corrected to account for fully aged material by subtracting approximately 100 mV from values calculated for the base metal. The shift in Ecritical (threshold potential 1) also appears to be approximately 100 mV in most cases. Thus, the difference Ecritical-Ecorr appears to be virtually unchanged. The effect of thermal aging on the corrosion rate is accounted for in the enhancement factor, Gaged, and is based upon a ratio of the non-equilibrium current densities for base metal and aged material. The value of Gaged for base metal is approximately one (Gaged ~ 1) whereas the value of Gaged for fully aged material is larger (Gaged ~ 2.5). Material with less precipitation than the fully aged material would have an intermediate value of Gaged (1 = Gaged = 2.5). Assume that Gaged is uniformly distributed between these limits and that this distribution is half uncertainty and half variability. ANL-EBS-MD-000003 REV 00 103 January 2000 6.8 MICROBIAL INFLUENCED CORROSION It has been observed that nickel-based alloys such as Alloy 22 are relatively resistant to microbial influenced corrosion (Lian et al. 1999). Furthermore, it is believed that microbial growth in the repository will be limited by the availability of nutrients. For example, H+ is known to be generated by bacterial isolates from Yucca Mountain. Furthermore, thiobaccilus ferro-oxidans oxidize Fe2+, while geobacter metallireducens reduce Fe3+. Other microbes can reduce SO4 2- and produce S2-. Ultimately, the impact of MIC will be accounted for by adjusting Ecorr, Ecritical, pH, and the sulfide concentration. The possible acceleration of abiotic corrosion processes by microbial growth is addressed here. Horn (1999) has shown that MIC can enhance corrosion rates of Alloy 22 by a factor of at least two. Measurements for Alloy 22 and other similar materials are shown in Table 25. Figure 47 is a schematic representation of the corrosion model for the Alloy 22 outer barrier. The augmentation of corrosion rates due to MIC are accounted for in the model as shown in Figure 48; here GMIC is the enhancement factor. effective MIC effective dt dp G dt dp × = (Eq. 55) This factor is calculated as the ratio of corrosion rates (microbes to sterile) and from Table 25. The value of GMIC for Alloy 22 in sterile media is approximately one (GMIC ~ 1), whereas the value of GMIC for Alloy 22 in inoculated media is larger (GMIC ~ 2). Assume that GMIC is uniformly distributed between these limits and that this distribution is half uncertainty and half variability. A patch experiencing both thermal aging an MIC would have a corrosion rate enhanced by the factor Gaged × GMIC. The principal nutrient-limiting factor to microbial growth in situ at Yucca Mountain has been determined to be low levels of phosphate. There is virtually no phosphate contained in J-13 groundwater. Yucca Mountain bacteria grown in the presence of Yucca Mountain tuff are apparently able to solubilize phosphate contained in the tuff to support growth to levels of 106 cells ml-1 of groundwater. When exogenous phosphate is added (10 mM), the levels of bacterial growth increase to 107 to 108 cells ml-1. The one to two orders-of-magnitude difference in bacterial growth with and without the presence of exogenous phosphate is almost certainly not significant with respect to effects on corrosion rates. Therefore, nutrient limitation, at least at a first approximation, was not factored into the overall MIC model. It may be noted, however, that the two-fold GMIC included in the model was in the presence of sufficient phosphate to sustain higher levels of bacterial growth (in an effort to achieve accelerated conditions). Other environmental factors that could effect levels of bacterial growth include temperature and radiation. These factors, however, are closely coupled to RH; as temperature and radiation decrease in the repository, RH is predicted to increase. At the same time, while there are some types of microorganisms that can survive elevated temperatures (= 120oC) and high radiation doses if there is no available water, then bacterial activity is completely prevented. Thus, because water availability is the primary limiting factor and this factor is coupled to other less critical limiting factors, water availability (as expressed by RH) was used as the primary gauge of microbial activity. ANL-EBS-MD-000003 REV 00 104 January 2000 Determination of a critical mass of total bacteria required to cause MIC is not an issue that needs to be addressed in the MIC model. Bacterial densities in Yucca Mountain rock have been determined to be on the order of 104 to 105 cells gm-1 of rock. In absolute terms, this is almost certainly above the threshold required to cause MIC. Further, bacterial densities were shown to increase one to two orders-of-magnitude when water is available (above). A more germane concern is the types of bacteria present, their abundance, and how their relative numbers are affected when water is available for growth. Corrosion rates will be affected (at least on some WP materials) for example, if organic acid producers out compete sulfate reducers or inorganic acid producers for available nutrients when water is sufficient to support growth. No data is currently available regarding the composition of the bacterial community over the changing environmental conditions anticipated during repository evolution. Instead, this issue has been addressed in the current model by determining overall corrosion rates under a standardized set of conditions, in the presence and absence of a defined set of characterized Yucca Mountain bacteria. Clearly, more data is required to better predict MIC on any given material with respect to this concern. Corrosion rates are currently being determined in the presence of Yucca Mountain rock containing the complete complement of Yucca Mountain bacteria and under conditions more representative of the repository. MIC is defined as a localized effect; thus, not all areas are equivalent on any given waste package with respect to bacterial colonization. It is well documented that bacteria preferentially colonize weldments, heat-affected zones, and charged regions (Borenstein and White 1989; Walsh 1989; Enos and Taylor 1996). However, the current model is based on data collected using unwelded specimens. In order to account for preferential areas of colonization in the model, it might be assumed that GMIC is uniformly distributed with respect to a real distribution. Table 25. Alterations in Corrosion Potentials Associated with Microbial Degradation Tested Sample Initial Condition Average Corrosion Rate (µm/yr) Corrosion Potential Ecorr (V vs. SCE) Initial Endpoint CS1020 + YM Microbes 8.8 -0.660 -0.685 Sterile CS 1020 1.4 -0.500 -0.550 M400 + YM Microbes 1.02 -0.415 -0.315 Sterile M400 0.005 -0.135 -0.070 C-22 + YM Microbes 0.022 -0.440 -0.252 Sterile C-22 0.011 -0.260 -0.200 I625 + YM Microbes 0.013 -0.440 -0.285 Sterile I625 0.003 -0.160 -0.130 304SS + YM Microbes 0.035 -0.540 -0.280 Sterile 304SS 0.003 -0.145 -0.065 DTN: LL991203505924.094 NOTE: Horn (1999) ANL-EBS-MD-000003 REV 00 105 January 2000 6.9 RECENTLY GENERATED DATA FOR ABSTRACTION AMRS 6.9.1 Two-Year LTCTF Data Rates of GC based upon 6- and 12-month exposures are discussed in Section 6.5.2. Very recently, data representing 24 months of exposure has become available. Those data are included in this section so that it can also be included in Abstraction AMRs and WAPDEG analyses. As previously discussed, tests in the LTCTF represent three generic water chemistries. SDW has 10× the ionic content of J-13 well water, while SCW has 1000× the ionic content. The measured pH levels of the 10× and 1000× J-13 well waters are 9.5 to 10. SAW is an acidified water that is around 4000× the ionic content of J-13 well water with a pH of approximately 2.7. Not all salts in the water will concentrate to these levels because of their limited solubilities, but the more soluble anions such as chloride, sulfate, and nitrate (which have the biggest effects on corrosion) will concentrate to these levels. Specimens are tested at two temperatures (60 and 90°C) for each of the three water chemistries. Half of the numbers of specimens are fully immersed in the water while the remaining half are exposed to the wet vapor above the water. A few specimens are also placed right at the water line so that their exposed area is half in the vapor, half in the water. Half of the numbers of test specimens contain welds. There were at least 144 test specimens measured during each exposure period. These general corrosion rates are obtained gravimetrically by the weight loss experienced during the exposure periods. The variation in measured general corrosion rates on Alloy 22 is decreasing with increased exposure time. The ranges of general corrosion rates measured at three time intervals (6-, 12-, and 24-months of exposure) are: 6-month exposure: range -0.06 to +0.73 µm y-1, mean 0.05 µm y-1 12-month exposure: range -.0.04 to +0.10 µm y-1, mean 0.03 µm y-1 24-month exposure: range -0.03 to +0.07 µm y-1, mean 0.01 µm y-1 Measurements on the order of 0.01 µm y-1 are around the experimental accuracy of this method. By far, the greatest variation in corrosion rates was measured in the first 6 months of exposure. The mean value of the corrosion rate after 24 months of exposure is 0.01 µm y-1. The corrosion rates do not appear to depend much at all on the temperature and chemical composition of the water tested thus far. Extrapolation of this mean value to 10,000 years would mean an average consumption of only 0.1 mm out of a thickness of 2 cm proposed for the Alloy 22 outer barrier of the waste package. Even at the highest rate measured in this data set, the maximum consumption would be less than 1 mm over the 10,000 year time period. Negative corrosion rates indicate a weight gain by the specimen even after all corrosion products and oxides from the surface have been thoroughly cleaned off. ANL-EBS-MD-000003 REV 00 106 January 2000 Cumulative distribution functions generated with 24-month data alone are shown in Figures 47 through 50. Cumulative distribution functions generated with a combined data set representing 6, 12- and 24-month data are shown in Figures 51 and 52. The curve shown in Figure 51 includes apparent negative rates, while those negative values have been eliminated from the curve shown in Figure 52. The curve shown in Figure 52 is summarized in Table 26. The distributions based upon the 24-month data are more narrow than comparable distributions based upon 6- and 12-month data. Since rates are calculated by dividing exposure time into the weight loss, a doubling of exposure time reduces the estimated error by a factor of two. While outliers were observed in the 6- and 12-month data, none were observed in the 24-month (two-year) data. It is believed that these more recent data will greatly alleviate the range of predicted failure times to times well beyond the period sought for compliance with the requirement of substantially complete containment. In observing the surfaces of the exposed specimens for all three time-periods, no evidence of LC has been observed. Specimens are mounted to the supporting test racks by Teflon® coated fasteners and washers. These washers create an intentional crevice to provide a surface area where crevice effects (electrolyte more concentrated than base solution). In addition, one type of specimen uses a special Teflon crevice former that is spring loaded to ensure that the contact is maintained between washer and specimen (crevice effects are more severe in tight crevices). Teflon has a tendency to creep at these test temperatures resulting in a looser crevice with the passage of time. Examination of plastically strained U-bend specimens, again for all three time periods, indicates no initiation of SCC in both the base material and in the welded material. Half the number of these U-bend specimens contained welds. The significance of the observations indicating no localized corrosion (that is no pits, no crevice attack, no intergranular attack) and no stress corrosion crack initiation, as well as a very low general corrosion rate, assures that Alloy 22 will provide an extremely long lived waste package. The longer these corrosion tests operate, the greater will be our assurance of the performance of this material. In the coming year, we plan to add another test environment in the long term corrosion test facility, an environment corresponding to ‘saturated’ conditions of water dripping, evaporating, and ionic salt concentrating on a hot metal surface. This environment will be more concentrated in chloride and nitrate (most soluble of the ionic species) and somewhat higher in pH than the solutions already under test. Short-term electrochemical tests already indicate that Alloy 22 does not corrode appreciably in this environment, but the longer term exposure test will be needed for confirmation of these results. Thus, the results from the two-year exposure period are very encouraging for Alloy 22. Compared to data generated from earlier exposure time periods, the most recent data set provides greater confidence of the projected corrosion performance of this material. ANL-EBS-MD-000003 REV 00 107 January 2000 0 10 20 30 40 50 60 70 80 90 100 -40 -30 -20 -10 0 10 20 30 40 Penetration Rate (nm/yr) Percentile (%) Two-Year LTCTF Data for Alloy 22 - Generic Weight Loss Samples DTN: LL000112205924.112 Figure 47. Additional Two-Year GC Corrosion Rate Data from LTCTF Based upon Generic Weight Loss Samples 0 10 20 30 40 50 60 70 80 90 100 -20 -10 0 10 20 30 40 50 60 70 80 Penetration Rate (nm/yr) Percentile (%) Two-Year LTCTF Data for Alloy 22 - Generic Crevice Samples DTN: LL000112205924.112 Figure 48. Additional Two-Year GC Corrosion Rate Data from LTCTF Based upon Generic Crevice Samples ANL-EBS-MD-000003 REV 00 108 January 2000 0 10 20 30 40 50 60 70 80 90 100 -80 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90 Penetration Rate (nm/yr) Percentile (%) Two-Year LTCTF Data for Alloy 22 - Generic Weight Loss and Crevice Samples Combined DTN: LL000112205924.112 Figure 49. Additional Two-Year GC Corrosion Rate Data from LTCTF Based upon Both Generic Weight Loss and Crevice Samples, including those with Apparent Negative Rates 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 Penetration Rate (nm/yr) Percentile (%) Two-Year LTCTF Data for Alloy 22 - Generic Weight Loss and Crevice Samples - No Negative Rates DTN: LL000112205924.112 Figure 50. Additional Two-Year GC Corrosion Rate Data from LTCTF Based upon Both Generic Weight Loss and Crevice samples, including those with Apparent Negative Rates ANL-EBS-MD-000003 REV 00 109 January 2000 0 10 20 30 40 50 60 70 80 90 100 -100 0 100 200 300 400 500 600 700 800 Penetration Rate (nm/yr) Percentile (%) Combination of All LTCTF GC Rate Data for Alloy 22 (6, 12 & 24 Month Exposures) DTN: LL000112205924.112 Figure 51. Combination of All GC Rate Data for Alloy 22 from LTCTF, including 6-, 12- and 24-Month (Two-Year) Exposures 0 10 20 30 40 50 60 70 80 90 100 0 100 200 300 400 500 600 700 800 Penetration Rate (nm/yr) Percentile (%) Combination of All LTCTF GC Rate Data for Alloy 22 (6, 12 & 24 Month Exposures) - No Negative Rates DTN: LL000112205924.112 Figure 52. Combination of All GC Rate Data for Alloy 22 from LTCTF, including 6-, 12- and 24-Month (Two-Year) Exposures with Negative Rates Excluded ANL-EBS-MD-000003 REV 00 110 January 2000 Table 26. Summary of the Distribution Shown in Figure 52 Percentile (%) Penetration Rate (nm y-1) 0.00 0 5.20 2.07 10.00 4.21 50.40 26.64 90.00 97.99 95.20 112.54 97.60 143.08 99.20 250.56 99.60 467.28 100.00 730.77 DTN: LL000112205924.112 6.9.2 Additional CP Data for BSW Test Media Several CP measurements have now been made with BSW electrolytes and are summarized in Table 27. The corresponding curves are shown in Figures 53 through 56. As previously discussed, extreme aging of Alloy 22 can shift the corrosion potential in a less noble (cathodic) direction by approximately 100 mV. This is accompanied by a slight increase in non-equilibrium passive current densities. There is some evidence of an anodic oxidation peak, characteristic of type 2 curves. For the present time, we will classify these CP curves as type 1-2. Table 27. Electrochemical Potentials Determined from CP Curves Sample ID Aging Time Aging Temp. Medium Temp. Reversal Potential Corrosio n Potential Threshold Potential 1 CP Curve Type hours oC oC mV mV mV DEA158 10 700 BSW 110oC 1200 -233 418 Type 1-2 DEA159 10 700 BSW 110oC 1200 -257 419 Type 1-2 DEA208 173 700 BSW 110oC 1200 -345 394 Type 1-2 DEA209 173 700 BSW 110oC 1200 -372 361 Type 1-2 DTN: LL000112105924.111 ANL-EBS-MD-000003 REV 00 111 January 2000 -600 -400 -200 0 200 400 600 800 1000 1200 1.E-09 1.E-08 1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 Current (A) Potential (mV vs Ag/AgCl) (DEA158) DTN: LL000112105924.111 Figure 53. CP Curve for Thermally Alloy 22 in 110°C BSW – Aged at 700°C for 10 Hours (DEA158) -600 -400 -200 0 200 400 600 800 1000 1200 1.E-09 1.E-08 1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 Current (A) Potential (mV vs Ag/AgCl) (DEA159) DTN: LL000112105924.111 Figure 54. CP Curve for Thermally Alloy 22 in 110°C BSW – Aged at 700°C for 10 Hours (DEA159) ANL-EBS-MD-000003 REV 00 112 January 2000 -600 -400 -200 0 200 400 600 800 1000 1200 1.E-09 1.E-08 1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 Current (A) Potential (mV vs Ag/AgCl) (DEA208) DTN: LL000112105924.111 Figure 55. CP Curve for Thermally Alloy 22 in 110°C BSW – Aged at 700°C for 173 Hours (DEA208) -600 -400 -200 0 200 400 600 800 1000 1200 1.E-09 1.E-08 1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 Current (A) Potential (mV vs Ag/AgCl) (DEA209) DTN: LL000112105924.111 Figure 56. CP Curve for Thermally Alloy 22 in 110°C BSW – Aged at 700°C for 173 Hours (DEA209) ANL-EBS-MD-000003 REV 00 113 January 2000 6.10 SUMMARY OF MODEL The model for the general and localized corrosion of Alloy 22 is summarized in Figures 57 and 58. The threshold RH is first used to determine whether or not DOX will take place. If DOX is determined to occur, the parabolic growth law represented by Equations 11 and 13 is then used to calculate the corrosion rate as a function of temperature. If the threshold RH is exceeded, HAC will occur in the absence of dripping water, and APC will occur in the presence of dripping water. If APC is assumed to occur, the corrosion and critical potentials are used to determine whether the mode of attack is general or localized. The correlations represented by Equation 17 and Table 5 can be used as the basis for estimating these potentials at the 50th percentile. Since the material specifications can be based partly on the measured corrosion and critical potentials, it can be assumed that these potentials will be uniformly distributed about the 50th percentile values determined from the correlation. For example, the 0th and 100th percentile values of Ecorr can be assumed to be at Ecorr (50th percentile) ± 75 mV. This acceptable margin was determined by splitting the differences shown in Table 6. Similarly, the 0th and 100th percentile values of Ecritical can be assumed to be at Ecritical (50th percentile) ± 75 mV. In principle, material falling outside of these specified ranges would not be accepted. Other equivalent correlations of Ecorr and Ecritical, based upon data relevant to the repository, can also be used. If the comparison of Ecorr to Ecritical indicates GC, the distribution of rates determined from the LTCTF will be used as the basis of the GC rate. A study of four test coupons of Alloy 22 removed from the LTCTF after one year showed varying degrees of coverage by silicate deposits but no evidence of localized corrosion by pitting. The distributions of GC rate shown in Sections 6.5.2 and 6.5.4 can be corrected for the maximum bias due to SiO2 deposit formation by adding a constant value of 63 nm y-1 to each estimated value of the GC rate. This is equivalent to shifting the curves shown in Figures 23, 25 and 26 to the right by 63 nm y-1. If the comparison indicates localized corrosion, the distribution of rates presented in Table 22 will be used. Corrosion rates will be enhanced to account for MIC above 90% RH. The effect of thermal aging on the corrosion rate is accounted for in the enhancement factor, Gaged, and is based upon a ratio of the non-equilibrium current densities for base metal and aged material. The value of Gaged for base metal is approximately one (Gaged ~ 1), whereas the value of Gaged for fully aged material is larger (Gaged ~ 2.5). Material with less precipitation than the fully aged material would have an intermediate value of Gaged (1 = Gaged = 2.5). ANL-EBS-MD-000003 REV 00 114 January 2000 Figure 57. Schematic Representation of Corrosion Model for Alloy 22 Outer Barrier critical RH RH = ? Dripping DOX dt dp HAC dt dp GC dt dp LC dt dp critical corr E E = C T ° = 100 ) ( ) ( ) ( 3 2 1 T f i T f E T f E SCW pass critical corr = = = ) ( ) ( ) ( 6 5 4 T f i T f E T f E SSW pass critical corr = = = Dripping RH T , , Effective dt dp yes yes yes yes no no no no ANL-EBS-MD-000003 REV 00 115 January 2000 Figure 58. Schematic Representation Showing Augmentation of Model to Account for MIC % 90 = RH 1 = MIC G Effective MIC Effective dt dp G dt dp × = yes no Effective dt dp 1 = MIC G ANL-EBS-MD-000003 REV 00 116 January 2000 INTENTIONALLY LEFT BLANK ANL-EBS-MD-000003 REV 00 117 January 2000 7. CONCLUSIONS Alloy 22 is an extremely Corrosion Resistant Material, with a very stable passive film. Based upon exposures in the LTCTF, the GC rates of Alloy 22 are typically below the level of detection, with four outliers having reported rates up to 0.75 µm per year. In any event, over the 10,000 year life of the repository, GC of the Alloy 22 (assumed to be 2 cm thick) should not be life limiting. Because measured corrosion potentials are far below threshold potentials, localized breakdown of the passive film is unlikely under plausible conditions, even in SSW at 120°C. The pH in ambient-temperature crevices formed from Alloy 22 have been determined experimentally, with only modest lowering of the crevice pH observed under plausible conditions. Extreme lowering of the crevice pH was only observed under situations where the applied potential at the crevice mouth was sufficient to result in catastrophic breakdown of the passive film above the threshold potential in non-buffered conditions not characteristic of the Yucca Mountain environment. In cases where naturally occurring buffers are present in the crevice solution, little or no lowering of the pH was observed, even with significant applied potential. With exposures of twelve months, no evidence of crevice corrosion has been observed in SDW, SCW, and SAW at temperatures up to 90°C. An abstracted model has been presented, with parameters determined experimentally, that should enable performance assessment to account for the general and localized corrosion of this material. A feature of this model is the use of the materials specification to limit the range of corrosion and threshold potentials, thereby making sure that substandard materials prone to localized attack are avoided. Model validation will be covered in part by a companion AMR on abstraction of this model. This document and its conclusions may be affected by technical product input information that requires confirmation. Any changes to the document or its conclusions that may occur as a result of completing the confirmation activities will be reflected in subsequent revisions. The status of the input information quality may be confirmed by review of the Document Input Reference System database. As examples, the status of AFM results shown here will have little impact on quantitative results, as the data is only corroborative and any MIC or aging results could impact GC rates by a factor of four. ANL-EBS-MD-000003 REV 00 118 January 2000 INTENTIONALLY LEFT BLANK ANL-EBS-MD-000003 REV 00 119 January 2000 8. REFERENCES Andresen, P.L. 198. "Modeling of Water and Material Chemistry Effects on Crack Tip Chemistry and Resulting Crack Growth Kinetics." Proceedings of the Third International Symposium on Environmental Degradation of Materials in Nuclear Power Systems-Water Reactors held August 30-September 3, 1987 in Traverse City, Michigan. 301-312. Warrendale, Pennsylvania: The Metallurgical Society. TIC: 246608. Asphahani, A.I. 1980. "Corrosion Resistance of High Performance Alloys." Materials Performance, 19, (12), 33-43. Houston, Texas: National Association of Corrosion Engineers. TIC: 240562. ASTM (American Society for Testing and Materials) 1987. ASTM G1-81. 1987. Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens. Philadelphia, Pennsylvania: American Society for Testing and Materials. TIC: 246089. ASTM 1989. ASTM G5-87. 1989. Standard Reference Test Method for Making Potentiostatic and Potentiodynamic Anodic Polarization Measurements. Philadelphia, Pennsylvania: American Society for Testing and Materials. TIC: 246088. ASTM 1997a. ASTM B 575-94. 1994. Standard Specification for Low-Carbon Nickel- Molybdenum-Chromium and Low-Carbon Nickel-Chromium-Molybdenum Steel Alloy Plate, Sheet, and Strip. Philadelphia, Pennsylvania: American Society for Testing and Materials. TIC: 237683. ASTM 1997b. Not used. ASTM 1997c. Not used. ASTM 1997d. ASTM G 5 - 94. Standard Reference Test Method for Making Potentiostatic and Potentiodynamic Anodic Polarization Measurements. Philadelphia, Pennsylvania: American Society for Testing and Materials. TIC: 231902. ASTM 1997e. ASTM G 1-90 (Reapproved 1999). 1990. Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens. West Conshohocken, Pennsylvania: American Society for Testing and Materials. TIC: 238771. Bard, A.J. and Faulkner, L.R. 1980. Electrochemical Methods, Fundamentals and Applications. 62-72. New York, New York: John Wiley & Sons. TIC: 245183. Bedrossian, P. 1999. High-Resolution Measurements of Aqueous and Environmental Corrosion and Oxidation Processes Using Photolithography and Atomic Force Microscopy. AP-E-20-70. Livermore, California: Lawrence Livermore National Laboratory. ACC: MOL.19991203.0258. Borenstein, S.W. and White, D.C. 1989. "Influence of Welding Variables on Microbiologically Influenced Corrosion of Austenitic Stainless Steel Weldments." Corrosion 89 April-17-21, 1989 ANL-EBS-MD-000003 REV 00 120 January 2000 New Orleans Convention Center, New Orleans, Louisiana, Paper Number 183, 183/1 - 183/13. Houston, Texas: National Association of Corrosion Engineers. TIC: 246583. Burr, I.W. 1974. "The Normal Curve." Applied Statistical Methods. 135-142. New York, New York: Academic Press. TIC: 245317. Cramer, S.D. 1974. "The Solubility of Oxygen in Geothermal Brines." Corrosion Problems in Energy Conversion and Generation. Tedmen, C.S., Jr., ed. 251-262. Princeton, New Jersey: Electrochemical Society. TIC: 245180. CRWMS M&O 1999a. License Application Design Selection Report. B00000000-01717-4600- 00123 REV 01 ICN 01. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.19990908.0319. CRWMS M&O 1999b. General Corrosion and Localized Corrosion of Waste Package Outer Barrier. Work Direction and Planning Document. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.19990708.0230. CRWMS M&O 1999c. Classification of the MGR Uncanistered Spent Nuclear Fuel Disposal Container System. ANL-UDC-SE-000001 REV 00. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.19990928.0216. CRWMS M&O 1999d. 11017040 Long Term Materials Testing and Modeling. Activity Evaluation. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.19990224.0429. CRWMS M&O 1999e. Waste Package Material Properties. BBA000000-01717-0210-00017 REV 00. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.19990407.0172. CRWMS M&O 1999f. Uncanistered Spent Nuclear Fuel Disposal Container System Description Document. SDD-UDC-SE-000001 REV 00. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.19991217.0512. CRWMS M&O 2000a. Input Transmittal for Environment on the Surfaces of the Drip Shield and Waste Package Outer Barrier. WP-LNL-00018.T. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.20000118.0316. CRWMS M&O 2000b. Input Transmittal for Aging and Phase Stability of Waste Package Outer Barrier Model. WP-LNL-00019.T. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.20000118.0317. CRWMS M&O 2000c. Input Transmittal for General Corrosion and Localized Corrosion of the Drip Shield. WP-LNL-00022T. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.20000118.0318. DOE (U.S. Department of Energy) 1998. Quality Assurance Requirements and Description. DOE/RW-0333P, Rev. 8. Washington, D.C.: U.S. Department of Energy, Office of Civilian Radioactive Waste Management. ACC: MOL.19980601.0022. ANL-EBS-MD-000003 REV 00 121 January 2000 Enos, D.G. and Taylor, S.R. 1996. "Influence of Sulfate-Reducing Bacteria on Alloy 625 and Austenitic Stainless Steel Weldments." Corrosion Science, 52, (11), 831-843. Houston, Texas: National Association of Corrosion Engineers International. TIC: 246554. Estill, J.C. 1998. "Long-Term Corrosion Studies." Engineered Materials Characterization Report. UCRL-ID-119564. Vol. 3. Livermore, California: Lawrence Livermore National Laboratory. ACC: MOL.19981222.0137. Evans, M.; Hastings, N.; and Peacock, B. 1993. Statistical Distributions. 2nd Edition. New York, New York: John Wiley & Sons. TIC: 246114. Farmer, J.C. and McCright, R.D. 1998. "Crevice Corrosion and Pitting of High-Level Waste Containers: Integration of Deterministic and Probabilistic Models." Proceedings of Corrosion 98, March 22-27, 1998, San Diego, California, 160/1 to 160/24. Houston, Texas: NACE International. TIC: 238172. Farmer, J.C.; McCright, R.D.; Estill, J.C.; and Gordon, S.R. 1998. Development of Integrated Mechanistically-Based Degradation-Mode Models for Performance Assessment of High-Level Waste Containers. URCL-JC-130811, Rev. 1. Livermore, California: Lawrence Livermore National Laboratory. ACC: MOL.19990729.0007. Farmer, J.C.; McCright, R.D.; Estill, J.C.; and Gordon, S.R. 1999. "Development of Integrated Mechanistically-Based Degradation-Mode Models for Performance Assessment of High-Level Waste Containers." Scientific Basis for Nuclear Waste Management XXII, Materials Research Society Symposium held November 30-December 4, 1998, Boston, Massachusetts, U.S.A., D.J. Wronkiewicz and J.H. Lee eds. 556, 855-862. Warrendale, Pennsylvania: Materials Research Society. TIC: 246426. Galvele, J.R. 1976. "Transport Processes and the Mechanism of Pitting of Metals." Journal Electrochemical Society, 123, (4), 464-474. New York, New York: Electrochemical Society. TIC: 239904. Gartland, P.O. 1997. "A Simple Model of Crevice Corrosion Propagation for Stainless Steels in Seawater." Corrosion 97, 417/1 to 417/17. Houston, Texas: NACE International. TIC: 245216. Gdowski, GE 1991. Survey of Degradation Modes of Four Nickel-Chromium-Molybdenum Alloys. UCRL-ID-108330. Livermore, California: Lawrence Livermore National Laboratory. ACC: NNA.19910521.0010. Gdowski, G.E. 1997a. Formulation and Make-Up of Simulated Dilute Water, Low Ionic Content Aqueous Solution. UCRL-ID-132285. Livermore, California: Lawrence Livermore National Laboratory. TIC: 245767. Gdowski, G.E. 1997b. Formulation and Make-Up of Simulated Concentrated Water(SCW), High Ionic Content Aqueous Solution. UCRL-ID-132286. Livermore, California: Lawrence Livermore National Laboratory. TIC: 245768. ANL-EBS-MD-000003 REV 00 122 January 2000 Gdowski, G.E. 1997c. . Formulation and Make-Up of Simulated Acidic Concentrated Water (SAW), High Ionic Content Aqueous Solution. UCRL-ID-132287. Livermore, California: Lawrence Livermore National Laboratory. TIC: 245766. Glass, R.S.; Overturf, G.E.; Van Konynenburg, R.A.; and McCright, R.D. 1986. "Gamma Radiation Effects on Corrosion-I. Electrochemical Mechanisms for the Aqueous Corrosion Processes of Austenitic Stainless Steels Relevant to Nuclear Waste Disposal in Tuff." Corrosion Science, 26, (8), 577-590. Oxford, Great Britain: Pergamon. TIC: 226179. Green, R.A. 1999. "Data Acquisition System Software Routines for Activity E-20-81, Development of Crevice Corrosion Models" Letter from Green, R.A. (LLNL) to Farmer, J., June, 17, 1999, CMS 99-006, with attachments ACC: MOL.20000125.0685. Gruss, K.A.; Cragnolino, G.A.; Dunn, D.S.; and Sridhar, N. 1998. "Repassivation Potential for Localized Corrosion of Alloys 625 and C22 in Simulated Repository Environments." Proceedings of Corrosion 98, March 22-27, 1998, San Diego, California, 149/1 to 149/15. Houston, Texas: NACE International. TIC: 237149. Hack, H.P. 1983. "Crevice Corrosion Behavior of Molybdenum-Containing Stainless Steels in Seawater." Materials Performance, 22, (6), 24-30. Houston, Texas: NACE International. TIC: 245826. Harrar, J.E.; Carley, J.F.; Isherwood, W.F.; and Raber, E. 1990. Report of the Committee to Review the Use of J-13 Well Water in Nevada Nuclear Waste Storage Investigations. UCID- 21867. Livermore, California: Lawrence Livermore National Laboratory. ACC: NNA.19910131.0274. Haynes International. 1988. Hastelloy Alloy C-22. Kokomo, Indiana: Haynes International. TIC: 239938. Haynes International. 1987. Hastelloy Alloy C-276. Kokomo, Indiana: Haynes International. TIC: 234999. Horn, J.A. 1999. Approach and Supporting Data for MIC Modeling. WP267DM5. Livermore, California: Lawrence Livermore National Laboratory. ACC: MOL.20000125.0687. Horn, J.M.; Rivera, A.; Lian, T.; and Jones, D. 1998. "MIC Evaluation and Testing for the Yucca Mountain Repository." Proceedings of Corrosion 98, National Association of Corrosion Engineers, March 22-27, San Diego, California, 152/2 to 152/14. Houston, Texas: NACE International. TIC: 237146. Jones, D.A. 1996. Principles and Prevention of Corrosion. 2nd Edition. Upper Saddle River, New Jersey: Prentice Hall. TIC: 241233. Kim, Y.J. 1987. "Effect of Gamma Radiation on Electrochemical Behavior of 9 Cr-1Mo Alloy in NaCl Solutions." Journal of the Corrosion Science Society of Korea, 16, (1), 25-30. Seoul, Korea: Corrosion Science Society of Korea. TIC: 246337. ANL-EBS-MD-000003 REV 00 123 January 2000 Kim, Y-K. 1988. "Electrochemical Mechanisms for Radiation Corrosion Processes of 316 Austenitic Stainless Steel in Chloride Environment." Journal of the Corrosion Science Society of Korea, 17, (1), 20-26. Seoul, Korea: Korean Corrosion Science Society. TIC: 246622. Kim, Y-J. 1999a. "Analysis of Oxide Film Formed on Type 304 Stainless Steel in 288[degrees]C Water Containing Oxygen, Hydrogen, and Hydrogen Peroxide." Corrosion, 55, (1), 81-88. Houston, Texas: NACE International. TIC: 245184. Kim, Y-J. 1999b. "In-Situ Electrochemical Impedance Measurement of Oxide Film on 304 SS in 288°C Water." Proceedings of Corrosion 99, April 25-30, 1999, San Antonio, Texas , 437/1 to 437/11. Houston, Texas: NACE International. TIC: 246024. Leygraf, C. 1995. "Atmospheric Corrosion." Chapter 12 of Corrosion Mechanisms in Theory and Practice. Marcus, P. and Oudar, J., eds. New York, New York: Marcel Dekker. TIC: 104415. Lian, T.; Martin, S.; Jones, D.; Rivera, A.; and Horn, J. 1999. "Corrosion of Candidate Container Materials by Yucca Mountain Bacteria." Corrosion 99, Paper No. 476, Houston, Texas: NACE International. TIC: 245833 LL990610105924.074. Cyclic Polarization Calculations. Submittal date: 06/12/1999. LL990610205924.075. Cyclic Polarization Calculations with CPDATA2C. Submittal date: 06/12/1999. LL990610305924.076. Crevice Corrosion Data from Scientific Notebook SN051999. Submittal date: 06/12/1999. LL990610505924.078. Crevice Corrosion Data for 316L and C-22. Submittal date: 06/13/99. LL990610605924.079. LTCTF Data for C-22, TIGR7, TIGR12 and TIGR16. Submittal date: 06/13/1999. LL991203505924.094. Approach and Supporting Data for MIC Modeling. Submittal date: 12/13/1999. ACC: MOL.20000128.0142 LL991208205924.096. Deliquescence Point for Sodium Nitrate Solutions. Submittal date: 12/20/1999. ACC: MOL.20000128.0145 LL991208505924.099. GC Rates of Alloy 22. Submittal date: 12/20/1999. ACC: MOL.20000128.0140 LL991208605924.100. Coefficients for the Correlation of Crevice pH with Applied Potential. Submittal date: 12/20/1999. ACC: MOL.20000128.0139 LL991208705924.101. Effect of Thermal Aging on the Corrosion Resistance of Alloy 22. Submittal date: 12/20/99. ACC: MOL.20000128.0138 ANL-EBS-MD-000003 REV 00 124 January 2000 LL991213805924.110. BSW Recipes. Submittal date: 01/03/2000. ACC: MOL.20000128.0137. LL000112105924.111. Corrosion and Potential Data Associated with High pH Saturated Water. Submittal date: 01/25/2000. LL000112205924.112. BSW Water Data. Submittal date: 01/25/2000. Newman, J.S. 1991. "Infinitely Dilute Solutions." Electrochemical Systems, 2nd Edition. Section 11.5. 250-251. Englewood Cliffs, New Jersey: Prentice Hall. TIC: 245901. NRC 1999. Issue Resolution Status Report Key Technical Issue: Container Life and Source Term Rev. 2. Washington, D.C.: U.S. Nuclear Regulatory Commission. TIC: 245538. Nystrom, E.A.; Lee, J.B.; Sagues, A.A.; and Pickering, H.W. 1994. "An Approach for Estimating Anodic Current Distributions in Crevice Corrosion from Potential Measurements." Journal of the Electrochemical Society, 141, (2), 358-361. Pennington, New Jersey: Electrochemical Society. TIC: 236472. Oldfield, J.W. and Sutton, W.H. 1978. "Crevice Corrosion of Stainless Steels. I. A Mathematical Model." British Corrosion Journal, 13, (1), 13-22. London, England: Institute of Materials. TIC: 212461. Pickering, H.W. and Frankenthal, R.P. 1972. "On the Mechanism of Localized Corrosion of Iron and Stainless Steel. I. Electrochemical Studies." Journal of the Electrochemical Society, 119, (10), 1297-1304. Pennington, New Jersey: Journal of the Electrochemical Society. TIC: 245828. Scully, J.R.; Hudson, J.L.; Lunt, T.; Ilevbare, G.; and Kehler, B. 1999. Localized Corrosion Initiation and Transition to Stabilization in Alloys 625 and C-22. 1-27. Charlottesville, Virginia: University of Virginia. TIC: 246630. Sedriks, A.J. 1996. Corrosion of Stainless Steels. 2nd Edition. 179, 377. New York, New York: John Wiley & Sons. TIC: 245121. Sridhar, N. and Dunn, D.S. 1994. "Effect of Applied Potential on Changes in Solution Chemistry Inside Crevices on Type 304L Stainless Steel and Alloy 825." Corrosion, 50, (11), 857-872. Houston, Texas: National Association of Corrosion Engineers International. TIC: 245832. Treseder, R.S.; Baboian, R.; and Munger, C.G. 1991. NACE Corrosion Engineer's Reference Book. 156-181. Houston, Texas: NACE International. TIC: 245834. Walsh, D.W. 1999. "The Effects of Microstructure on MIC Susceptibility in High Strength Aluminum Alloys." Corrosion 99, Paper 187, 1-13. Houston, Texas: National Association of Corrosion Engineers. TIC: 246549. Walton, J.C.; Cragnolino, G.; and Kalandros, S.K. 1996. "A Numerical Model of Crevice Corrosion for Passive and Active Metals." Corrosion Science, 38, (1), 1-18. Amsterdam, The Netherlands: Pergamon. TIC: 233439. ANL-EBS-MD-000003 REV 00 125 January 2000 Weast, R.C., ed. 1974. CRC Handbook of Chemistry and Physics. 59th Edition. Page B-161. West Palm Beach, Florida: CRC Press. TIC: 246619. Welsch, G. and Pramod, D.D. eds. 1996. Oxidation and Corrosion of Intermetallic Alloys. West Layfayette, Indiana: Purdue University. TIC: 245280. ANL-EBS-MD-000003 REV 00 126 January 2000 INTENTIONALLY LEFT BLANK ANL-EBS-MD-000003 REV 00 127 January 2000 ATTACHMENTS Attachments to this document are listed below. The CD Rom includes a data inventory in the form of an Excel spreadsheet. Attachment Title I Report on AFM Study II Data Inventory Sheet ANL-EBS-MD-000003 REV 00 128 January 2000 INTENTIONALLY LEFT BLANK ANL-EBS-MD-000003 REV 00 I-1 of 38 January 2000 ATTACHMENT I REPORT ON AFM STUDY BY P. J. BEDROSSIAN ANL-EBS-MD-000003 REV 00 I-2 of 38 January 2000 INTENTIONALLY LEFT BLANK ANL-EBS-MD-000003 REV 00 I-3 of 38 January 2000 Surface Topographies of One-Year Weight-Loss Coupons of Alloy C-22™ from Long-Term Corrosion Testing Peter J. Bedrossian Division of Materials Science & Technology Lawrence Livermore National Laboratory, Livermore CA 94551 11 June 1999 1. ABSTRACT An atomic force microscope (AFM) to characterize the surface topographies of weight-loss coupons of Alloy C-22™ which had been exposed to two different environments in the Long- Term Corrosion Test Facility (LTCTF) at LLNL has been used for one year. A silicate deposit on these coupons, with the most extensive coverage occurring on the coupon immersed in an acidified bath has been observed. Localized corrosion on these coupons has not been detected. 2. INTRODUCTION The LTCTF at LLNL is an array of tanks holding various aqueous baths with controlled electrolyte concentrations at 60 or 90°C, in which coupons of candidate materials for the Waste Package are held in either aqueous (below the water line) or vapor (above the water line) phase conditions and removed periodically for analysis. Although the LTCTF coupons have primarily been used for analysis of general corrosion via weight loss, the objective of the present study has been the search for signs of localized corrosion, if any. The “weight loss” coupons are 2 inches long, 1 inch wide, and 1/8 inch thick. Descriptions of the LTCTF and its uses, along with the detailed composition of the aqueous environments, are contained in Reference [1]. The AFM, with sub-nanometer vertical resolution, is an ideal tool for detecting pit initiation in localized areas. We have applied AFM to five “weight loss” coupons of Alloy C-22™: one control coupon that was never in any bath (DWA163), one aqueous phase sample from a simulated acidified well water SAW (DWA051), one vapor phase sample from SAW (DWA048), one aqueous phase sample from a simulated alkaline concentrated water SCW (DWA120) and one vapor phase sample from SCW (DWA117). ANL-EBS-MD-000003 REV 00 I-4 of 38 January 2000 3. RESULTS AND DISCUSSION Representative AFM data are collected and displayed below. Each set of data consists of a largearea scan of at least 25x25 µm followed by smaller-area details of the region displayed in the large-area scan. We have used a Digital Instruments DM3100 AFM. After the coupons were removed from the LTCTF, they were ultrasonically agitated in deionized water, acetone, and methanol for ten minutes each. In general, the gross surface topography of the weight-loss coupons is dominated by the machining grooves, with typical heights of several hundred nanometers and typical lateral periodicities of several microns. The machining features on a bare surface are plainly visible on the images of coupon DWA163. Those samples which were removed from the LTCTF exhibit varying degrees of coverage by a deposit on top of this gross topography. X-ray diffraction scans of all five coupons show that the deposit is predominantly a silicate or SiO2, with some NaCl appearing on the two samples which were exposed in the SAW tank. The AFM images show that the most extensive coverage of the deposit occurred on test coupon DWA051, which was immersed in the SAW bath. The next most extensive coverage occurred on test coupon DWA048, which was held above the water line in the SAW bath. The two test coupons removed from the SCW bath showed lesser degrees of coverage by the silicate deposit in both the AFM images and the X-ray diffraction scans. Incomplete surface coverage by the silicate deposit often results in the appearance of surface depressions, particularly on the DWA051 coupon. Data collected to date do not show any of these depressions extending below the metal surface, because the bottoms of the holes are typically flat. One illustration of the analysis leading to this conclusion is shown below in the profile measured along the line trace marked in the image pb990607.023, which spans two such holes. As shown in the profile, the bottoms of the holes are flat, as would be expected for an interruption that occurs only in the silicate deposit. The following data are presented in this attachment, with page numbers listed. ANL-EBS-MD-000003 REV 00 I-5 of 38 January 2000 FIGURES Page 1. Control Coupon DWA163 pb990607.019 AFM Image........................................... I-7 2. Control Coupon DWA163 pb990607.020 AFM Image........................................... I-8 3. Control Coupon DWA163 pb990607.021 AFM Image........................................... I-9 4. Control Coupon DWA163 pb990607.022 AFM Image......................................... I-10 5. SAW Test Coupons: X-Ray Spectra of scales on SAW Coupons ........................ I-11 6. SAW, 90°C, Aqueous DWA051 pb990607.023 AFM Image ............................... I-12 7. SAW, 90°C, Aqueous DWA051 pb990607.023 AFM Image, top view ............... I-13 8. SAW, 90°C, Aqueous DWA051 Line Profile in pb990607.023 AFM Image ....... I-14 9. SAW, 90°C, Aqueous DWA051 pb990607.024 AFM Image ............................... I-15 10. SAW, 90°C, Aqueous DWA051 pb990607.033 AFM Image ............................... I-16 11. SAW, 90°C, Aqueous DWA051 pb990607.033 AFM Image ............................... I-17 12. SAW, 90°C, Aqueous DWA051 pb990607.029 AFM Image ............................... I-18 13. SAW, 90°C, Aqueous DWA051 pb990607.030 AFM Image ............................... I-19 14. SAW, 90°C, Aqueous DWA051 pb990607.031 AFM Image ............................... I-20 15. SAW, 90°C, Aqueous DWA051 pb990607.032 AFM Image ............................... I-21 16. SAW, 90°C, Aqueous DWA051 pb990607.032 AFM Image ............................... I-22 17. SAW, 90°C, Vapor DWA048 pb990607.046 AFM Image.................................... I-23 18. SAW, 90°C, Vapor DWA048 pb990607.045 AFM Image.................................... I-24 19. SAW, 90°C, Vapor DWA048 pb990607.048 AFM Image.................................... I-25 20. SAW, 90°C, Vapor DWA048 pb990607.050 AFM Image.................................... I-26 21. SAW, 90°C, Vapor DWA048 pb990607.054 AFM Image.................................... I-27 22. SCW Test Coupons: X-Ray Spectra of Scales on SCW Test Coupons ................. I-28 23. SCW, 90°C, Aqueous DWA 120 pb990607.001 AFM Image............................... I-29 24. SCW, 90°C, Aqueous DWA 120 pb990607.005 AFM Image............................... I-30 25. SCW, 90°C, Aqueous DWA 120 pb990607.015 AFM Image............................... I-31 26. SCW, 90°C, Vapor DWA117 pb990607.039 AFM Image.................................... I-32 27. SCW, 90°C, Vapor DWA117 pb990607.035 AFM Image.................................... I-33 28. SCW, 90°C, Vapor DWA117 pb990607.037 AFM Image.................................... I-34 29. SCW, 90°C, Vapor DWA117 pb990607.044 AFM Image.................................... I-35 30. SCW, 90°C, Vapor DWA117 pb990607.041 AFM Image.................................... I-36 31. SCW, 90°C, Vapor DWA117 pb990607.042 AFM Image.................................... I-37 ANL-EBS-MD-000003 REV 00 I-6 of 38 January 2000 4. SUMMARY A study of four test coupons of Alloy C-22™ removed from the LTCTF after one year showed varying degrees of coverage by silicate deposits but no evidence of localized corrosion by pitting. 5. ACKNOWLEDGMENTS The author is grateful to David Fix for his extensive AFM data collection, to Dominic Delgiudice for the x-ray measurements, to John Estill and Kenneth King for providing samples from the LTCTF, and to Joseph Farmer, Daniel McCright, and Ronald Musket for helpful discussions. This work was conducted at LLNL under the auspices of the US Department of Energy under Contract W-7405-Eng-48, and was supported by the Yucca Mountain Program. 1 J C Farmer, et al., “Development of Integrated Meghanistically-Based Degradation-Mode Models for Performance Assessment of High-Level Waste Containers,” UCRL-ID-130811 (1998), pp. 3 and 49 (Farmer et al. 1998). ANL-EBS-MD-000003 REV 00 I-7 of 38 January 2000 Figure 1. Control Coupon DWA163 pb990607.019 AFM Image Machining Grooves ANL-EBS-MD-000003 REV 00 I-8 of 38 January 2000 Figure 2. Control Coupon DWA163 pb990607.020 AFM Image Machining Grooves ANL-EBS-MD-000003 REV 00 I-9 of 38 January 2000 Figure 3. Control Coupon DWA163 pb990607.021 AFM Image ANL-EBS-MD-000003 REV 00 I-10 of 38 January 2000 Figure 4. Control Coupon DWA163 pb990607.022 AFM Image ANL-EBS-MD-000003 REV 00 I-11 of 38 January 2000 800 600 400 200 0 36 34 32 30 28 26 24 22 20 2-Theta (deg) SiO2 SiO2 NaCl DWA163 (spectrum z03309) DWA051 (spectrum z03320) DWA048 (spectrum z03316) Figure 5. SAW Test Coupons: X-Ray Spectra of scales on SAW Coupons ANL-EBS-MD-000003 REV 00 I-12 of 38 January 2000 Figure 6. SAW, 90°C, Aqueous DWA051 pb990607.023 AFM Image Deposit ANL-EBS-MD-000003 REV 00 I-13 of 38 January 2000 Figure 7. SAW, 90°C, Aqueous DWA051 pb990607.023 AFM Image, top view ANL-EBS-MD-000003 REV 00 I-14 of 38 January 2000 Figure 8. SAW, 90°C, Aqueous DWA051 Line Profile in pb990607.023 AFM Image ANL-EBS-MD-000003 REV 00 I-15 of 38 January 2000 Figure 9. SAW, 90°C, Aqueous DWA051 pb990607.024 AFM Image Hole in Deposit ANL-EBS-MD-000003 REV 00 I-16 of 38 January 2000 Figure 10. SAW, 90°C, Aqueous DWA051 pb990607.033 AFM Image ANL-EBS-MD-000003 REV 00 I-17 of 38 January 2000 Figure 11. SAW, 90°C, Aqueous DWA051 pb990607.033 AFM Image ANL-EBS-MD-000003 REV 00 I-18 of 38 January 2000 Figure 12. SAW, 90°C, Aqueous DWA051 pb990607.029 AFM Image ANL-EBS-MD-000003 REV 00 I-19 of 38 January 2000 Figure 13. SAW, 90°C, Aqueous DWA051 pb990607.030 AFM Image ANL-EBS-MD-000003 REV 00 I-20 of 38 January 2000 Figure 14. SAW, 90°C, Aqueous DWA051 pb990607.031 AFM Image ANL-EBS-MD-000003 REV 00 I-21 of 38 January 2000 Figure 15. SAW, 90°C, Aqueous DWA051 pb990607.032 AFM Image ANL-EBS-MD-000003 REV 00 I-22 of 38 January 2000 Figure 16. SAW, 90°C, Aqueous DWA051 pb990607.032 AFM Image ANL-EBS-MD-000003 REV 00 I-23 of 38 January 2000 Figure 17. SAW, 90°C, Vapor DWA048 pb990607.046 AFM Image ANL-EBS-MD-000003 REV 00 I-24 of 38 January 2000 Figure 18. SAW, 90°C, Vapor DWA048 pb990607.045 AFM Image ANL-EBS-MD-000003 REV 00 I-25 of 38 January 2000 Figure 19. SAW, 90°C, Vapor DWA048 pb990607.048 AFM Image ANL-EBS-MD-000003 REV 00 I-26 of 38 January 2000 Figure 20. SAW, 90°C, Vapor DWA048 pb990607.050 AFM Image ANL-EBS-MD-000003 REV 00 I-27 of 38 January 2000 Figure 21. SAW, 90°C, Vapor DWA048 pb990607.054 AFM Image Deposits ANL-EBS-MD-000003 REV 00 I-28 of 38 January 2000 500 400 300 200 100 0 36 34 32 30 28 26 24 22 20 2-Theta (deg) SiO2 DWA163 (spectrum z03309) DWB120 (spectrum z03762) DWA117 (spectrum z03763) Figure 22. SCW Test Coupons: X-Ray Spectra of Scales on SCW Test Coupons ANL-EBS-MD-000003 REV 00 I-29 of 38 January 2000 Figure 23. SCW, 90°C, Aqueous DWA 120 pb990607.001 AFM Image ANL-EBS-MD-000003 REV 00 I-30 of 38 January 2000 Figure 24. SCW, 90°C, Aqueous DWA 120 pb990607.005 AFM Image ANL-EBS-MD-000003 REV 00 I-31 of 38 January 2000 Figure 25. SCW, 90°C, Aqueous DWA 120 pb990607.015 AFM Image ANL-EBS-MD-000003 REV 00 I-32 of 38 January 2000 Figure 26. SCW, 90°C, Vapor DWA117 pb990607.039 AFM Image ANL-EBS-MD-000003 REV 00 I-33 of 38 January 2000 Figure 27. SCW, 90°C, Vapor DWA117 pb990607.035 AFM Image ANL-EBS-MD-000003 REV 00 I-34 of 38 January 2000 Figure 28. SCW, 90°C, Vapor DWA117 pb990607.037 AFM Image ANL-EBS-MD-000003 REV 00 I-35 of 38 January 2000 Figure 29. SCW, 90°C, Vapor DWA117 pb990607.044 AFM Image ANL-EBS-MD-000003 REV 00 I-36 of 38 January 2000 Figure 30. SCW, 90°C, Vapor DWA117 pb990607.041 AFM Image ANL-EBS-MD-000003 REV 00 I-37 of 38 January 2000 Figure 31. SCW, 90°C, Vapor DWA117 pb990607.042 AFM Image ANL-EBS-MD-000003 REV 00 I-38 of 38 January 2000 INTENTIONALLY LEFT BLANK ANL-EBS-MD-000003 REV 00 II-1 of 128 January 2000 ATTACHMENT II INVENTORY OF METAL SAMPLES WITH TRACEABILITY ANL-EBS-MD-000003 REV 00 II-2 of 128 January 2000 INTENTIONALLY LEFT BLANK