Hanford Double Shell Waste Tank Corrosion Studies (Final Report FY2019)

Abstract

VCIs formulations were added during mid-course of experiments, i.e., after coupons have experienced corrosion in the untreated GW simulant. Three tests were conducted using VCI-A and VCI-B. The first two tests were conducted using 100% recommended dosages of VCI-A and VCI-B. The third test was conducted at 10% of the recommended dosage of VCI-B. Following conclusions are made from the experimental data and results:  The corrosion rate data indicated that 10% of the recommended dosage is insufficient in mitigating corrosion. This observation is consistent with a prior study which also concluded that VCIs' effectiveness vanishes at 10% of the recommended dosages for the aboveground tank bottom underside application.  The data also showed that 100% recommended dosages of VCI-A and VCI-B mitigated pitting corrosion of weathered coupons. Specifically, VCI-A mitigated pitting corrosion in immersed, Level 2, and Level 3 coupons, whereas VCI-B mitigated pitting corrosion in immersed, Level 1, and Level 3 coupons. Statistical significance of corrosion rate decrease in Level 1 coupons for 100% recommended dosage of VCI-A and Level 2 coupons for 100% recommended dosage of VCI-B could not be established; this may be due to choice of coupons' surface orientation being vertical during the tests, leading to limited and uneven weathering during groundwater only and groundwater plus VCI exposures. 3. Long term OCP Drift Tests OCPs of carbon steel were measured in three simulants from the new limits testing. Two sets of coupons, with three coupons in each set, were fabricated with differing surface conditions. The coupons' surface conditions included a 600-grit polished surface, and a surface with mill-scale plus corrosion products from legacy carbon steel (rail car steel AAR TC-128). Two coupons, one coupon from each set, were placed in each chemistry, and OCPs of the coupons were monitored for five months. OCPs of the coupons with polished surfaces evolved in the anodic direction with respect to the initial values, whereas OCPs of two of the three coupons with mill-scale plus corrosion products evolved in cathodic direction with respect to the initial values. Terminal OCP of one mill-scale plus corrosion products coupon was about 100 mV anodic with respect to the initial value. CPP data of the polished coupons before OCP evolutions showed mixed responses, i.e., a clear delineation between pitting and no-pitting cannot be made, and CPP data after OCP evolutions remained mixed (category 3), indicating that change in OCP values did not affect the CPP responses. Following exposures of the coupons during OCP evolutions, CPP data for the mill-scale plus corrosion products coupons in the three simulants showed negative hysteresis (category 1), which is a sign of no-pitting. The OCP and CPP data indicate that simulant chemistry and surface condition affect extent and direction of OCP evolution, however, OCP evolution does not affect the pitting susceptibility. Electrochemical Impedance Spectroscopy (EIS) spectra of the polished coupon...

Chemistry Envelope for Pitting and Stress Corrosion Cracking Mitigation

“…SRNL has conducted laboratory tests to investigate pitting corrosion at current and anticipated DST waste chemistries for the past 5 years [9][10][11][12][13]. The testing utilized a protocol for cyclic potentiodynamic polarization (CPP), developed in conjunction with the Tank Integrity Expert Panel Corrosion Sub-group (TIEP-CSG), to assess the susceptibility of carbon steel to pitting corrosion in DST chemistries [14].…”

“…The modified ASTM G192 method is based on the Tsujikawa-Hisamatsu Electrochemical (THE) method and can be used to determine the repassivation potential. In this case, it was used to provide a definite categorization of pass or fail for borderline cases that showed mixed hysteresis [9][10][11][12][13]. If the results were a pass, "0" was used and if it was a fail "1" was used.…”

“…Logistic regression analysis was performed to develop the model for pitting susceptibility prediction [11][12][13]. Logistic regression assumes a binary response, pass or fail in this instance.…”

“…The coefficients for the species variables generated from the probability model were utilized in the development of an empirical "pitting factor" to provide a criterion for pitting susceptibility [12,13]. Equation 4shows the final form of the pitting factor.…”

“…This initial envelope was a compromise with the present corrosion chemistry requirements, which allow a minimum hydroxide concentration of 0.01 M hydroxide. However, a review of this proposed envelope during FY19 challenged the use of the higher probability region between 0.01 M and 0.1 M hydroxide [13]. A review of historical pitting data also demonstrated that there was a lack of data in this region.…”