Repeatability Limits and Reproducibility Limits

Among numerous quantitative analytical instruments, including Optical Emission Spectrometer, XRF fluorescence spectrometers, carbon and sulfur analyzers, oxygen, nitrogen, and hydrogen analyzers, ICP, and atomic absorption spectrometers, determining the reliability and compliance of test results is crucial. Therefore, repeatability limits and reproducibility limits are two key indicators, primarily used to define the acceptable range of error.

I. Concepts:

1. Repeatability Limit

- Conditions:

Same person, same instrument, same laboratory, short time period, same sample

- Purpose:

To measure the stability of the method/instrument/personnel

To determine whether the parallel sample is acceptable

Difference between two results > Repeatability Limit → Data is unreliable and needs to be repeated

2. Reproducibility Limit

- Conditions:

 Different people, different instruments, different laboratories, different time periods, same sample

- Purpose:

To measure the universality of the method in different scenarios

To determine whether the results from different laboratories are consistent

Difference between results from two laboratories > Reproducibility Limit → Results are not comparable

II. Standards:

1. Repeatability Limit

Under repeatability conditions, the absolute difference between two independent test results should not exceed the repeatability limit (r) listed in Table 8-15. The number of cases exceeding the repeatability limit (r) should not exceed 5%.

Table 8-15 Repeatability Limits (r) for Copper and Copper Alloy Samples

2. Reproducibility Limit

Under reproducibility conditions, the absolute difference between two independent test results should not exceed the reproducibility limit (R) listed in Table 8-16. The number of cases exceeding the reproducibility limit (R) should not exceed 5%.

Table 8-16 Reproducibility Limits (R) of Copper and Copper Alloy Samples