How Insurance Works

Vanta™ GeoChem Technical Tutorial Series Part 8: Quality Assurance and Quality Control


hi I’m Todd Houlihan welcome to the eighth in our video series on best practice use of portable XRF for giacomo applications I’m joined by Alex Thurston today our applications engineering manager how you doing Alex doing well Todd thanks okay this video is all about Quality Assurance quality control QA QC our customs are starting to generate their own chemistry on their samples so they need to take responsibility for the integrity of that data right and those users that frequently send samples to the lab are likely to have already a an established QA QC procedure so some of the aspects of this video may not be entirely new to them however I do think that they will pick up on some unique aspects of handheld XRF my customers always ask what QA QC should they do so I always reply to them well why don’t you replicate what you’re doing for the lab that way they create uniformity they can standardize their protocols and on a basic level it’s just easier for them to remember right I think it boils down to essentially three aspects that would be contamination accuracy and repeatability okay the XRF contamination check this is where we use a blank sample to test the cleanliness of the front of the analyzer the window it also allows us to check if there’s any dust affecting the results we supply a silicon dioxide fused quartz disk with every analyzer and when measuring that we should only measure silicon right and the user should insert this blank at regular intervals during their batch testing right so the next thing to do is to check the accuracy of the analyzer so the customer could use all sorts of samples to do this either some matrix match certified reference materials some samples from their project that have been essayed by the lab or the NIST sample that we supply with every gear cam unit right and a good rule of thumb for the frequency of this type testing would be about every one in 20 tests ideally the accuracy level is up to user but it should be ultimately fit for their purpose right and that accuracy is going to be determined by a whole series of things that we’ve discussed in this series sample prep test times calibration the sample containers that we use so the user should send off a percentage of the samples that they’re testing with the vanta to a lab to confirm the analysis of the instrument and what that helps them do is build up a library of samples that they can ensure the accuracy of the vance to their needs is confirmed with the the known good lab analysis right more experiences they send off a lot of samples right at the beginning of a project but once they build up confidence in their method their workflows get some data back from the lab that confirms the accuracy they then drop the amount of samples that they send to the lab over the duration of a project so it’s hard could you speak to how we would check the XRF precision sure checking the XRF precision is all about testing the stability of the analyzer it’s a good way to check repeatability so multiple measurements on the same sample and looking at the variability of the data it’s more important than accuracy most of the time because we can use samples of known concentrations and adjust the instruments factory calibration to get the results that we need right and the precision of these results are critical to the reliability of that data and that’s something that we’ve worked on here in the factory in order to achieve that precision is by improving the hardware of the instrument as well as through a number of signal processing improvements that we call axon technology right and I’ve really seen that bear out when I do my relative standard deviation checks but precision measurements where we do a minimum of seven readings on the same sample and calculating the RS DS we can really see that our esteem really low right and the stability of the instrument is where the bands are really shines from working from weeks to months in harsh conditions high sample throughputs and in hot temperatures you can see on screen here that the stability of the band is maintained through temperature cycling and customers can check this precision really easily and the stability of their analyzer over time by just going back and monitoring those XRF accuracy checks that we talked about earlier if they look at that data over a long period of time it’ll help them identify if there’s anything out of spec and the the example we can see on the screen here was perfect because we can see that the anomaly is an anomaly identified associated with the instrument being out of spec after a certain period of time and Merrill out that allowed them to go back and check that I also like to do a lot of repeatability measurements of the same sample under different sample preparation regimes to help me optimize what sample prep is going to be best to achieve a customer’s data quality objectives I also like to do repeat measurements to define instrument error from sample error can you speak a little bit more about that yeah by taking multiple measurements of a sample but in different locations of that sample we get a better feeling for the heterogeneity of the sample and how much that is affecting the results if we test that same sample but leaving the analyzer in the same location then we define instrument error and by comparing those two allows us to optimize a sample prep that is going to work best to achieve certain data quality objectives that’s helpful and users can always contact us for more information on that subject right for sure yes so let’s summarize qa/qc we need to test blanks test some samples of known concentrations send those samples to the lab and do some repeatability measurements duplicates calculate some RS DS and for frequency a good rule of thumb is one in 20 or we can just repeat what we do for our laboratory procedure brilliant thanks Alex happy to help Todd next time we’ll be covering some tips on implementing your portable XRF program things we haven’t covered in the previous episodes so see you next time you


Leave a Reply

Your email address will not be published. Required fields are marked *