Validation of onsite salt analysis
Hello everyone, it has been a while since I have posted anything mainly due to preparation for an AIB audit and learning to implement the SQF code. After reading over the general requirements, I developed two questions that don't seem to have answers. One, how do we validate in house testing on finished products for salt (NaCl). I would think that an independent standard being ran weekly would suffice for this parameter. In my old job as an analytical chemist, I had a protocol that I followed such as; ICB, ICV, CCB, CCV, spikes, spike duplicates etc. However, in our plant, we are not certified by any agency on salt testing. Secondly, a question of when a lot number is or is not valid. In the past we have used the day a batch was started even if it carried over to the next day. However, I believe that, technically speaking, anything ran after 12:00 at night should be a different lot with the reasoning being that although the ingredients were batched the day before they were not processed until the next day. The key here being "processed" rather than simply weighed and put on a pallet. So, two different questions submitted for the forum to consider today. Thanks for the help.
Hello everyone, it has been a while since I have posted anything mainly due to preparation for an AIB audit and learning to implement the SQF code. After reading over the general requirements, I developed two questions that don't seem to have answers. One, how do we validate in house testing on finished products for salt (NaCl). I would think that an independent standard being ran weekly would suffice for this parameter. In my old job as an analytical chemist, I had a protocol that I followed such as; ICB, ICV, CCB, CCV, spikes, spike duplicates etc. However, in our plant, we are not certified by any agency on salt testing. Secondly, a question of when a lot number is or is not valid. In the past we have used the day a batch was started even if it carried over to the next day. However, I believe that, technically speaking, anything ran after 12:00 at night should be a different lot with the reasoning being that although the ingredients were batched the day before they were not processed until the next day. The key here being "processed" rather than simply weighed and put on a pallet. So, two different questions submitted for the forum to consider today. Thanks for the help.
Hi Taste maker,
I'll try the first one.
You omitted to mention what procedure is used for the NaCl measurement. And for whom the Validation is intended. AIB ? SQF ?
A full-blown validation can presumably be quite heavy-going using official procedures, eg -
http://www.ncbi.nlm....les/PMC3374569/
But i doubt such is required for the above organizations ?
Standard procedure / small ring test ? ++
Personally, for micro. I have typically followed offical procedures like AOAC/BAM and sent samples to 2-3 labs to compare data for quantitative things like APC. Accuracy for numerical micro. data is (fortunately) well established as poor.
Nobody has ever asked me to officially validate the "standard" titration method i use for chloride ion. From long memory NaCl is anyway a primary standard for AgNO3 :smile:
I assume you do not wish to achieve ISO 17025 ?
Yeah, I should have included the method as well in my question. I am using the AgNO3 titration using a G20 Compact Analyzer using a selective ion for Cl-. I am wondering if a third party chloride standard in the middle of the linear range of the instrument would suffice to verify the analyzer is working properly.
Hi Taste Maker,
You can validate your inhouse testing by following ways -
1) Using a published validated method of testing like AOAC
2) Chemist validation by getting the same lot tested by your chemist and an accredited laboratory. I am not sure of the acceptable standard deviation in your country. Most of the times if it is a bulk analysis, %RSD of 2% is usually acceptable. Prepare a schedule of such analysis. The data can be used to validate your in-house results.
I think it should work ! It has worked for us...not BRC but for other auditor's similar questions !
Regards