The data are from a benzene analyzer, measuring percentage volume of benzene in gasoline. The measurement that is displayed in the plot is [(analyzer feedback) - (a known standard sample value)], which centers the numbers about zero. Due to the very low variability shown (relative to the repeatability of the analyzer), there is some "jumpiness" to the data.

There are quite a few points in the first 3/4 of the period that were above zero, indicating that we had a "run rule" violation (more than seven consecutive points above or below the line). At the end of the data run, this situation was corrected, with the instrument calibration adjusted. Here is the analysis log:

• 01/31/2001 -- Being worked by Casillas, Stewart, and Adams. Instrument has been calibration checked a couple of times and is still above the mean. South and Offline charts are both OK.
• 02/07/2001 -- Checked calibration again - still OK. Changed out calibration standard - also OK.
• 02/08/2001 -- Re-ran calibration on both North and South - still OK.
• 02/11/2001 -- Continue to watch. The South chart and our GC SQC chart both look good. Casillas, Stewart, and Adams are looking into the problem.
• 02/15/2001 -- Note downward trend, but still above the mean. Will continue to watch.
• 02/20/2001 -- Finally dipped below the mean.