'1A' is an error character which is used when a character cannot be translated.
Is your example data the result of a TRANSLATE WITH ERROR? Your screenshot is showing ascii/latin but no unicode chars.
If that ColA is unicode OREPACE will not work, because it's defined with CHARACTER SET LATIN and there will be an automatic translation of ColA to unicode *before* the replace. And you can't use the latin character '1A' because it's the error char:
"Note: 0x1A in LATIN/KANJI1/KANJISJIS and U+FFFD in UNICODE/GRAPHIC are used internally by Teradata as the error character; therefore, they are unusable as user data. The user cannot store or retrieve these values through Teradata."
You might find out about the actual non-translatable chars using TRANSLATE_CHK and CHAR2HEXINT. If it's just a single char you could remove it using SUBSTRING.