[daf44]

Hi there,

For one of my projects, i need to determine the fat concentration (mg/l) in a water stream.
I've collected samples and send them to a lab where they can measure this.
I've asked them what the standard deviation of their analyze method is, but they responded that they don't know. Therefore, I would like to create an validation test for their analyze method.

My thoughts are to make a 1L solution of XX mg of fat (let's say 100 mg fat), divide this into ten 100 ml samples. These ten samples should then be measured to see if they all produce the same result (100 mg/l). The problem I foresee is that you divide the 1L solution into ten 100 ml samples, the concentration of fat won't be the same in each sample, because fat does not dissolve into water and therefore won't distribute evenly over each 100 ml sample.

What would be the best way to tackle this problem?

FYI: the analyze method consists of adding a surplus of acid to the water samples, after which the fat precipitates. They then seperate, dry and weigh the precipitation.

[Gern]

Hi,

First of all, an accredited lab should know what the reproducibility and repeatability of their analyses is. From this the SD could be calculated but the latter two parameters are of more interest.
Validation of the analytical methods really is the responsibility of the lab.

But, to come back to your question, I think it could be done in this way;
Dissolve the fat in an organic solvent which could be evaporated easily e.g chloroform. Mix the right amount of this solution with 100 ml of  water and evaporate the solvent a low temperature to prevent water evaporation. This should result in having 100 ml samples of water with 10 gm of fat. As fat is poorly soluble in water, the sample will not be homogeneous, so the lab should use the entire sample for the analysis.

Please be careful when handling chloroform, work in a fume hood and take all necessary after precautions like goggles, gloves, lab coat etc.

good luck