ABSTRACT
The production of formation sand is a problem associated with most oil deposits in the world. Major Sand production effects affect safety, well or field economics and continuous production. The ability to predict when a formation will fail and produce sand forms the basis as to what type of sand management strategy to use (whether down hole sand control system will be required or a sand management approach and that is why this research work is important as it provides a means of forecasting the production potential of a well more accurately than the conventional laboratory testing.
Three criteria have been identified as crucial in sand production. Formation strength, production rate and in-situ stresses. The modeling of the formation failure mechanism, which is linked to these three characteristics, is required for sand prediction. The expected failure mechanism is largely determined by the type of formation (consolidated, friable, or unconsolidated). All information on the structure is gathered at this point. Offset well data (production data, drilling, completion, and so on), geological information, coring, rock characteristics, and logging data, for example, are all examples. Data on in-situ field stresses has been gathered; formation density assessment causes overload stress or by applying a 1.12 psi/ft gradient to the depth in question.
In order to predict, this method uses a correlation between sand production well data and field operational factors. To evaluate the sand production potential and define a standard for sanding or no sanding, one or a collection of metrics is typically utilized. This is owing to the practical difficulties of monitoring and documenting data for all of the wells in a study for several years. Porosity, drawdown or flow rate, and other parameters are frequently utilized.
In conclusion the study states the importance of data accuracy in sand prediction.
At the end of this research work, the model successfully forecast quantitatively the rate of sand production in the tested wells, with a maximum deviation of less than 8% in all cases except one, which is below the model's specified tolerance of 10% deviation.
