Download: Blending study

Laboratory Projects can be completed to support the refinery blending process and enhance the overall accuracy of targeting the lubricity specification. A major issue is the validity of a batch result that fails but is near to the specification maximum. A refinery may require the batch to be re-blended with a higher dosage of additive and then re-tested. Alternatively the batch may be re-sampled and tested and the average result taken for batch release. Re- blending a batch can be costly disrupting planned schedules and such expense is best avoided if caused by random errors in a test procedure.

As a matter of routine, batch samples that fail or are borderline can be re-tested by a second laboratory and confidence taken to a level that such batches can be released without going through an expensive re-blending process. Historical records enable confidence to be established for a refinery process that can be used to overrule single test results.

The Blending Study.xlsm download above gives two example studies where a number of batch samples have been measured above the 460 microns maximum specification. In the first study, the random testing errors are shown to be the main factor and the HFRR target probably set only marginally high. In the second set of data, batches are proven to be out of specification and in need of re-blending before release in to the market.