-
- Posts: 9
- Joined: Mon Feb 11, 2013 5:50 pm
???
These units monitor processes which are usually very stable and measure around a certain point. Accuracy is desired around this point so that the analyzer can track the output it is monitoring precisely, as a trend over time. It is the trend not the exact measurement that is important.
For example, an ethylene stream may be measuring acetylene content at 0-5 ppm but it may calibrate at 1 ppm and when the trend reaches 3 ppm an alarm may go off. It is not necessary that the measurement may be off by 0.1 ppm at 3 ppm, but if the calibration is at 1 ppm and the measurement hits 3 ppm (which might actually be 2.9 or 3.1 ppm) then the unit sends its alarm and it has done its job.
The plant engineer has designed these factors into the plant. To actually calibrate over a large range of a measurement, say 5% 50% and 95% is unwise, as the cost of all these calibration standards is high, they require a lot of time to monitor their consumption and use, and they may not have a long lifetime where the value they were made does not change. Buy a 95% std today and 3 months from now it may be 88%, so why bother with all the expense and trouble? It gives more opportunities for mistakes to enter into the equation. Ooops, I used the 50% calibration instead of the 95% cal std, etc.
I hope I have given you a least a glimpse into the thinking of using only one cal std on an analyzer. Of course, if several streams are being monitored with a large difference in center of their ranges, then more than one cal std might be needed.
best wishes,
Rod
