AssaySynthesis

An AssaySynthesis document allows you to determine the complete description of a crude oil, given measurement data ranging in scope from just two or three measurements, through to complete crude assays. Assay intelligently uses its knowledge of crude oil behaviour to generate the crude description most consistent with the input data, flagging up and reduce the weight of any values which lie outside the range of the statistical model. The generated crude contains complete information, for all properties, and may be added to the database and subsequently manipulated just like any standard Database crude.

To use AssaySynthesis documents,

1. Click the button at the bottom of the Document Navigator and choose AssaySynthesis from the menu.

2. Click the New button in the Document group of the Home ribbon tab and choose AssaySynthesis from the menu.

Once an AssaySynthesis document is created you can set up the individual pages of the Document as necessary. After Generation you will be able to view the Profile Graphs.

The most suitable set of values to enter into a crude generation module in Assay will depend upon which values in the final generated crude are of most concern to evaluate accurately.

For a given property, crude oil property profiles are, generally, similar from crude to crude. In practice this means that in order to produce a very good assay synthesis, no more than a handful of values will generally be needed for any given property. Furthermore, Assay's strong inter-property correlations may mean it is already accurately predicting, say, viscosity profiles from entered density data. To the level of precision required for the application of that data, any further accuracy may not be needed, and hence it may not be necessary to enter viscosity data.

If possible, an assessment of the use to which the crude description will be put is valuable in allowing a sound determination of how many and which values should be used in an AssaySynthesis, or AssayUpdate. Too much data will not be a problem, if the data is consistent. The likelihood of rogue measurements increases as more values are used. Rogue measurements are intelligently given lower statistical importance in the generation procedure, but nevertheless contribute to the final generated crude description, leading potentially to less certain predictions (as indicated by the error lines).

• Three or four measurements for a given property will usually constrain its profile to required tolerances.

• Use the error lines as a guide to whether you have achieved the required tolerance for the property values of interest.

• Use the intelligent correlative methods of Assay to your advantage by considering groups of related properties together.