[quote=losttheplot]If you’re to take a sample from a bulk liquid and the solution is homogenous are AQL/RQL sampling plans valid (ie sampling plans such as n=9 a=0). If the solution is homogenous then the only variation in the 9 samples would be test method.
What sort of sampling plan could be applied - would a single sample suffice?
I would suggest that if you can prove that the bulk is homogenous, then only 1 sample will do. It also depends on what type of parameter you are testing in the bulk. if the parameter is a pass / fail or yes / no, then, if you can prove homogeneity then only one sample will do. If however you are testing variable parameters, %conc, visc., density, etc…then I think that want to see the variability to help in determining long term and short term process capability.