Gas Lift Application and Severe Slugging in Toe Down Horizontal Well

Gas lift is one of the most common artificial techniques employed in alleviating liquid loading problems in horizontal wells. To study the relative effects of several completion parameters, namely, end of tubing location, presence of a packer, injection location and injection rate, a large-scale test facility was constructed. This large-scale test facility also enabled studying the effect of pipe diameter on horizontal well flow behavior. Throughout the course of this study, the facility was set to a 1° toe down configuration.;In the first experimental phase, the flow characteristics in a 6 in. ID toe down tubingless horizontal well was studied. The onset of film reversal and initiation of liquid loading was experimentally determined and compared with Brito (2015) results, which was conducted on a 2 in. ID small scale facility allowing the analysis of the pipe diameter. Slug characteristics such as frequency and slug length were experimentally determined. It was observed that existing slug models in literature did not predict the observed slug characteristics accurately. Holdup and pressure characteristics were also studied and compared with Brito (2015).;In the second experimental phase, gas lift studies were conducted to determine the effects of the parameters mentioned earlier. A comprehensive experimental program was developed for this purpose. A 2 in. ID tubing was installed into the system for this experimental phase. Experimental insights led to the formulation of certain guidelines regarding the application of gas lift in toe down horizontal wells.;The third experimental phase was an exploratory study, which led to the discovery of the existence of severe slugging in toe down configurations. Severe slugging has thus far never been documented in such configurations. The experimental study investigated the effects of end of tubing location, liquid and gas rates, the presence of a packer and end of tubing eccentricity on this phenomenon.;Based on observations from the third experimental phase, a transient severe slugging prediction model was developed. This model was based on Balino et al. (2010) and was modified to account for slug flow in the lateral section and the cross-sectional area change associated with a toe down well. In addition, a single point steady state criterion similar to Boe (1981) was also developed for prediction of severe slugging in toe down wells.