How can you tell if chemical analyses are dependable

The question sometimes arises about how a client can be totally confident about the services supplied by a chemical analyst. This is very fair question and one that we ourselves need to consider if we outsource particular tests for which we do not have in-house facilities. The following provides a good example.

We wanted to know how much ethanol was present in a number of samples of four-star petrol. To gain confidence in the lab we used to do the tests, we took some petrol and added a precise amount of extra ethanol to it and included that as one of the samples. Taking another drop of the same petrol, we added even more ethanol to that and included that as a sample.

This is a method known by chemists as “spiking” and what it means is that although we didn’t know how much ethanol was in the original petrol, when the analyses came back, we should be able relate the results between the originals and the spiked samples.

Outcome; the results came back and were such that we could tell they were significantly in error. So then we had contact the lab, tell about the problem and see what they had to say. Fairly quickly, the Manager of the lab came back and admitted that the member of his staff who had done the work had developed a little “short cut” for the test method and used it without authorisation. He had another member of staff repeat the tests and this time the results made sense.

How common this type of problem is is difficult to say, but use of a the spiking technique, and various other self-checking methods you can use when using a lab will often enable you to find out who you can trust…and who you cannot!

Chemistry and Materials FAQsPermalink

Comments are closed.