When performing an accuracy check using a primary standard a question that will often arise is what concentration range would be considered acceptable?
Hach can offer some guidelines to be used to help identify or manage expectations for acceptance ranges.
When determining an acceptance range it is important to consider that error in a result is the sum of all individual error sources. Here are some examples of error sources and explainations of how some documentation regarding each can be helpful but should not be used alone to determine the acceptance range.
All prepared primary standards that are sold by Hach are certified. The lot specific Certificate of Analysis (COA) for the primary standard can be downloaded from the website (See: How can a Certificate of Analysis (CoA) for chemicals be obtained?
). The COA will often include a specification range which is the acceptance range for the standard that is used during quality control. The COA also often includes the result of the quality control test to show that at the time the quality test was performed, that lot of standard was within the specification range. This does not mean that the standard analyzed by an end user would fall within the specification range (See: How should the Certificate Of Analysis (COA) for standard solutions be interpreted?
). This specification range is for the standard only and does not take into consideration any other sources of error. It also only applies to the standard at the time that the quality test was performed and not to the time that the standard is ran by an end user.
The instrument and method:
Almost all method documents include a method performance section which often states something similar to "The method performance data that follows was derived from laboratory tests during ideal test conditions. Users can get different results under different test conditions." before providing a precision statement for the method. In summary, assuming ideal test conditions and that this method document is followed exactly when performing the test, the error in the result at that part of the range is expected to be within the tolerance provided with 95% confidence. (See: Can Hach advise on how deviations from the method procedure will impact results?
). Because this precision range is assuming ideal test conditions, this could be interpreted as the error that can be expected from the method and instrument as error sources. This does not however include any other error souces aside from the instrument and method, and therefore should not be used as the acceptance range for running a primary standard. If the primary standard that is being used is at a different concentration than the concentration the precision statement is specific to, a different tolerance could be expected as well. The percent error from the instrument and chemistry of the method would be expected to increase as the concentration approaches the estimated detection limit.
Other sources of error:
A list of other error sources to consider when could include but is not limited to:
Where in the measurement range the concentration of the primary standard is
accuracy of volume measurements
variations of reagent lot/reagent blank
sample cell condition
errors caused in dilution of the primary standard from sources like the glassware and dilution water quality
In conclusion the specification range from the COA and precision range from the method performance section of the method document can be useful in helping to determine the error from the standard, instrument, and method chemistry as error sources, but neither alone should be used to determine the acceptance range when running a primary standard as an accuracy check. When determining what would be considered an acceptable range when running a primary standard, all error sources in that reading should be considered.