Summary This paper provides insights on the design of downhole gas separators based on the laboratory study of various separator designs. A downhole gas separator, also known as a gas anchor, may be installed below the pump to separate free gas from the produced liquid. The free gas-produced downhole is usually separated through the casing-tubing annulus (the casing-tubing annulus acts as a natural gravity separator) while the liquid is produced through the tubing. However, inefficient gas anchor designs are widespread and an acceptable guide for their optimum design does not currently exist. Laboratory testing of downhole gas separators has been ongoing since January 2005 at The University of Texas at Austin Petroleum Production Engineering Facility (UTAPPEF) using an instrumented full-scale model of a wellbore and separator constructed with clear acrylic pipe to visualize the fluid mechanics of the separation process. An air and water mixture is injected through the well's perforations. The air and water flow rate measurements are used to measure and define a performance plot of each separator design. The separator designs tested differed in entry-port configuration, size of dip tube, and the relative position of the separator-fluid entry ports with respect to the well's perforations. Based on the results of the tests, a new separator design that includes the effect of centrifugal forces to separate the gas and liquid phases was developed. The results show that, for the conditions in the laboratory, 100% separation was achieved whenever the entry ports were located 1- to 2-feet (ft) below the bottom-most casing perforation thereby dispelling the predominant industry-held opinion that more distance is required between the separator-fluid entry ports and the bottom-most casing perforation. Similarly, laboratory results equally show that an optimum dip tube length of 5.5 ft is sufficient for the optimal separation of free gas from the produced liquid by the separator. This clearly runs contrary to the accepted industry practice, as many industry models employed for the estimation of dip-tube length were found to over-estimate the dip-tube length, and this subsequently results in the undesirable increased pressure drop across the separator. Lastly, the entry port geometry does not appear to have a significant impact on the separator performance as long as sufficient flow area is present. The efficiency of all gravity-driven separators was limited by the liquid velocity inside the separator annulus. When the liquid velocity inside the separator averaged approximately 6 in. per second or less, an almost-to-complete gas separation was achieved. On the other hand, the centrifugal separator had a liquid capacity 70% greater than any of the gravity-driven, static-downhole gas separators.
Read full abstract