Hi,

At the company I work, we supply skids for the loading of tank trucks. One of our customers is having the following problem: They claim that they load a truck and them send it to their customer. When the truck arrives at the customer it's unloaded and there is an difference in volume that's too big".


My answer and assumptions:

"Assuming that both loading and unloading skids are calibrated, the sistematic errors are almost gone, therefore there's only "random error"( wich is not equal to the uncertainty, but i will assume to be the uncertainty). As per OIML R117 for loading and unloading of liquids we have an accuracy class of 0,3% (0,3% is only an example of the accuracy class, it could be another value depending on the application) at 95% confidence level, so the maximum error in the volume is E=0,3% ( 95% confidence level). If I load a true value of 10000 L (I really will never know the true value) my skid will show a volume from 99970 L to 10030 L, then when I unload it the volume shown at the unloading skid will be a value between 99970L and 10030L. So in the worst case i will measure a value of 10000-E at the loading skid and a value of 10000+E at the unloading skid, so the maximum difference in the volume between them is 2E. Therefore if the difference in volume they are informing is below 2E the measuring system is ok."

I would like to know if my assumption is correct or not?