In an effort to improve corrosion monitoring and control at the Hanford Site, an eight-channel electrochemical noise (EN) based corrosion monitoring system was designed, fabricated and installed into double-shell tank 241-AN-107 in September 1997. This system is a larger scale version of a prototype system installed in double-shell tank 241-AZ-101 in August 1996 and monitors eight three-electrode channels positioned at different elevations in the tank. The system is capable of detecting the onset of pitting and stress corrosion cracking should tank waste conditions change to allow these mechanisms to occur. No finding
was provided for troubleshooting and data analysis for nearly a year following installation. When these activities were finally permitted, it was found that most of the data collected was corrupted to some degree by an unknown source of electrical interference. The source of this disturbance has not yet been located. Data collected between 10/97 and 1/98 do not appear to have been affected by the disturbances and signal interference that plague much of the rest of the data. Both summary statistical data and raw data collected during this time period indicate that uniform corrosion is the dominant active corrosion mechanism in tank 241-AN-107. Uniform corrosion rates were calculated from EN data. Despite correction for differences in surface area, uniform corrosion rates calculated from noise data collected on small pin-type electrodes do not agree well with rates measured on larger C-ring type electrodes. Between 10/97 and 1/98, C-rings in the supernate showed an average uniform corrosion rate of approximately 7 mils per year (mpy). Pin electrodes in the supernate over the same time period showed a uniform corrosion rate of less than 1 mpy. Which value is correct and the source of the difference have not yet been determined.