Originally Posted by

**Kostya** I am looking for some help with a very basic problem. There are three devices used for measuring volume. Device A and Device B measure volumes separately that are then combined and measured by Device C. Device A and B measure with a relative uncertainty of +- 5% with a 90% confidence interval. Device C measures with a relative uncertainty of +- 4% with a 90% confidence interval. When comparing the results from A + B with C what is the acceptable deviation? How far off can A+B be from C while still within tolerance. Thanks for any help in advance.