Pressure Drop In Wells Producing Oil And Gas

Abstract

ABSTRACT A simple, mechanistically based scheme for the calculation of the pressuredrop in wells producing oil and gas in the single-phase liquid, bubble and slugflow patterns is described and checked with independent field data. The schemeis based on an identification of the flow pattern through a modification of theflow pattern map of Govier, Radford and Dunn and the application of themechanical energy balance in a form appropriate for the flow pattern assuggested by Govier and Aziz. Predictions for some 48 wells are compared withfield data and with the predictions of Orkiszewski, Duns and Ros, and Hagedornand Brown. The proposed method gives results at least as good as any of theothers, is more soundly based on the mechanism of flow and is independent ofthe data with which it is confirmed. A computer program for the method and a typical printout are given. INTRODUCTION Many methods have been proposed for the estimation of the pressure drop inwells which produce a mixture of oil and gas. These are reviewed in detail by Govier and Aziz. Most of the methods are strictly empirical, but for certain ofthe flow patterns which may be encountered methods based on the mechanism offlow may now be developed. This is especially true for the bubble and the slugflow patterns and, as will be discussed in a following paper, the annular-mistflow pattern encountered with gas-condensate wells. Orkiszewski has developed a pressure drop prediction scheme based on anidentification of the flow pattern and the application of selected mechanisticand empirical methods to the individual flow patterns. His original appraisalincluded, among others, the Poettmann and Carpenter and related correlations.The most accurate methods, the Duns and Ros and the Hagedorn and Brown werethen tested along with his own scheme against 148 well conditions.Orkiszewski's comparison showed that his method gave improved accuracy overthat of Duns and Ros and Hagedorn and Brown. Espanol et al confirmedthis with data from 44 different wells.