Uncertainty Analysis of Oil Well Flow Rate on the Basis of Differential Entropy

Abstract

Oil well production efficiency depends on the accuracy of the flow rate prediction. The electrical submersible pumps are selecting and the well production control is carrying out based on the predicted values of flow rate. Inaccurate prediction may cause limitations of well deliverability or inefficient pumping. The prediction accuracy of flow rate changes in time related to initial data uncertainty that causes deviations between calculated flow rate values and measured ones. To minimize operating costs the same pump selection and control methods are used for groups of wells operating under the same conditions. However, sometimes wells demonstrate very different behavior even under the same conditions. In these wells flow rate changes becomes unpredictable by the common methods and additional studies required for correct prediction. The problem of finding wells with unpredictable flow rates at the early operation stages is very important because their inefficiency can significantly increase in time without special operation methods. The article considers the method of finding wells with potentially unpredictable flow rate changes with use of the entropy concept. The main feature of this method is that it is appropriate for data of any distribution types with given probability density function. The article discusses the relation between the value of joint reduction in uncertainty obtained from entropies of calculated flow rates and measured ones for a single well and the deviations between these flow rates. The novelty of the article is that the joint reduction in uncertainty in calculated value of well rate when knowing measured well rate is proposed as the measure of the well flow rate predictability.