I also used the Z data. When I imported in into NT I found that there were instances of duplicate rows pertaining to the first minute of the day on a handful of days, which I just removed.
For 1 minute data there are over 800,000 rows so there would have to be hundreds, if not thousands of...