已发表: 05/05/2016
已发表: 05/05/2016
The standard models of liquid loading in gas wells relate the "onset of liquid loading" to the concept of "critical gas velocity". As a mechanistic multiphase-flow concept, the critical gas velocity would be the smallest velocity of the upward flowing gas still providing enough drag to rule out liquid accumulation in the wellbore. The richness of the related phenomena in the gas field, however, comes from the interaction of multiphase-flow in the well and in the reservoir. The concept of critical gas velocity (or rate) is in fact misleading, because steady state upward flow of a gas-liquid mixture can happen at arbitrary low gas velocities, if the inlet boundary condition is constant-rate type. In addition, the actual transient liquid accumulation processes are virtually impossible to reproduce under laboratory conditions. Therefore, transforming laboratory observations directly into onset prediction models – however often it is done – has not lead to major breakthrough in comparison to what the pioneering "critical gas rate" models could already offer. In contrast, our work focuses on improvements in determining the overall liquid-content in the wellbore, and in assessing the strongly related flowing bottomhole pressure. Since in typical gas wells we still have to rely on observations at the wellhead, we look for a method which is capable to detect the gradual increase of the overall liquid-content in the wellbore, way before the actual liquid loading phenomena become obvious. The approach presented here relies on an empirical correlation that was originally developed to predict the liquid hold-up for multiphase upward flow affected by partial flow reversal in long vertical pipes. In general, the correlation can be used as an alternative method for calculating the gravitational component of the pressure difference between wellhead and bottomhole. Regarding the other components (friction and acceleration related) we rely on accepted industry standards. We show that our method provides better accuracy for bottomhole pressure calculation, using published data for 78 vertical gas wells. Then we make some recommendations for applying the method for deviated well sections.