Study of the Potential for an Off-Bottom Dynamic Kill of a Gas Well Having an Underground Blowout

Abstract

Summary. This paper reports on an analysis that links expected reservoir performance with wellbore hydraulics. The dynamics of counterflow of the kill fluid and gas below the kill string can play a significant role in the expected behavior during a dynamic kill. The design concepts and techniques presented should be applicable to any well that requires the kill string to be run or snubbed into the well or one in which the string is a significant distance off bottom at the time of a blowout and cannot be lowered to bottom. Introduction Industry literature on the concept of a dynamic well kill describes introduction of a kill fluid to the flowing gas stream at or near the flow zone in the blowing well. In a workover operation on a sanded-up gas well with a hole in the tubing, however, it may become necessary to perform a dynamic kill at the point above the flow zone where the last sand bridge is washed out. Arco Oil and Gas Co. became interested in whether the dynamic-well-kill method could be used, when sand was being washed out of a gas well with a hole in the tubing, to maintain sufficient control of the well to prevent an underground blowout if the sand bridges did not extend to the production zone in the well. To study the principles involved, we selected an example well from a gas-recycling EOR project. Wells in this reservoir produce 40 API [0.825-g/cm3] -gravity oil with GOR's of 30,000 to 40,000 scf/bbl [5403 to 7205 std m3/m3]. The gas is separated and injected into the reservoir. The reservoir pressure is thus maintained at about 4,700 psi [32 407 kPa]. The reservoir and gas properties are known. The example well was completed with 2 7/8-in. [7.3-cm] tubing, as Fig. 1 shows. Table 1 lists this well's maximum flow potential (MFP). To explore the risks that could result from an underground blowout, we assumed that a hole in the 2 7/8-in. [7.3-cm] tubing existed at 250 ft [76 m] and that the exit pressure was constant at 150 psi [1034 kPa], which is about a 10-lbm/gal [1198-kg/m3] fracture gradient. The well was assumed to be sanded up, with the bottom of the last sand bridge above the flow zone. The workover procedure under consideration was to snub into the well with a 1 1/4-in. [3.175-cm] tubing string, wash the sand out of the tubing, and if necessary, dynamically kill the well with 9.5-lbm/gal [1138-kg/m3] NaCl. Because the last sand bridge was washed out during the cleaning operation, the well could be opened to its MFP of about 25 MMscf/D [708 × 103 std m3/d] and an underground blowout would occur through the hole in the 2 7/8-in. [7.3-cm] tubing. Flow rates of this magnitude could conceivably crater the ground below the snubbing unit. Therefore, this type of operation needed the potential to kill the well dynamically when the last sand bridge was washed out at various assumed depths above the flow zone. Assessment of Potential for a Dynamic Kill Lynch et al. described a method for analyzing circulating dynamic kills of a gas well with reservoir performance data, wellbore hydraulics, and a kill string near the wellbore flow zone. Wellbore-hydraulics performance curves (Fig. 2) are generated for different circulation rates of a given fluid and compared with the reservoir performance curve. Any wellbore-hydraulics curve for a given circulation rate that intersects or falls below the reservoir performance curve (Curve B) will reach a stable flow condition, with the fluid being gas lifted out of the well. The minimum circulation rate to achieve a dynamic kill has a performance curve that is totally above (not intersecting) the reservoir performance curve (Curve A). The same approach can be used to analyze the dynamic kill potential of a kill string at any position above the flow zone in a well. Fig. 1 is a schematic of the wellbore geometry used to model the dynamic kill with 1 1/4-in. [3.175-cm] tubing inside 2 7/8-in. [7.3-cm] tubing. A reservoir-simulation/wellbore-hydraulics computer program was based on Clark and Perkins' work. To calculate the wellbore hydraulics, a single-phase flow model was used in the 2 7/8-in. [7.3-cm] tubing below the 1 1/4-in. [3.175 tubing, and a two-phase flow model was used in the 1 1/4 × 2 7/8-in. [3.175 × 7.3-cm] annulus. The well production data in Table 1 and reservoir properties in Table 2 were entered into a reservoir-simulation computer program to determine the expected flow rate of gas that the reservoir would deliver at the bottom of the 2 7/8-in. [7.3-cm] tubing for a given pressure at the perforations. Fig. 3 and 4 show the calculated reservoir deliverability as a function of wellbore bottomhole pressure (BHP). Table 3 lists the flowing gas composition used. For purposes of determining hydraulics within the wellbore, gas viscosity and density were calculated as functions of temperature and pressure. The reservoir pressure is approximately 8.5 lbm/gal [1018 kg/m3] equivalent, so a 9.5-lbm/gal [1138-kg/m3] NaCl brine was selected as the kill fluid. The viscosity of the brine as a function of temperature was calculated. The example well was directionally drilled, which is typical for the area. Hydraulic-pressure drops are related to measured depths (MD's), but fluid beads are related to true vertical depths (TVD's). For the example well, the relationship between TVD and MD was included in the computer program to calculate the hydraulic pressure drops and fluid heads properly at any given MD. The following assumptions and simplifications were used in calculating the wellbore-hydraulics curves for different circulation rates and led to a conservative design. 1.The increased OD at the connection upsets on the 1 1/4-in. [3.175-cm] tubing was ignored. 2. All downhole completion equipment, such as gas-lift mandrels and subsurface safety valves, were assumed to have the same ID as the 2 7/8-in. [7.3-cm] tubing with no leakage. 3. Oil production was not associated with the gas when brine was being circulated to kill the well. 4. No friction reducer was added to the 9.5-lbm/gal [1.138-kg/m3] NaCl brine. 5.The 2 7/8-in. [7.3-cm] tubing was clear of sand bridges and other flow restrictions from the perforations to the bottom of the 1 1/4-in. [3.175-cm] kill string. 6.The reservoir pressure was 4,700 psi [32 407 kPa], and no gas had been produced from this sanded-up well in the preceding 2 years. 7.The maximum surface-working-pressure limit with the 1 1/4-in. [3.175-cm] workstring and snubbing-unit pump system was 10,000 psi [68 950 kPa]. 8.The change from single-phase gas flow to two-phase gas and brine flow occurred exactly at the end of the 1 1/4-in. [3.175-cm tubing. SPEDE P. 215⁁

New Dynamic Kill Procedure for Off-Bottom Blowout Wells Considering Counter-Current Flow of Kill Fluid

“…Second, a method for calculating ZNLF holdup based on the procedure originally proposed by Duncan 2 is presented, and third, a general step by step procedure to apply all these concepts to an off-bottom dynamic kill is defined. Finally the results of the full-scale experiments in a 2787-foot deep research well by Flores-Avila et al 3 that validate the procedure are discussed, as well as an application example based on a case presented by Gillespie et al 4 .…”

Emergence of Delamination Fractures Around the Casing and Its Stability

“…Moreover, compared to other kinds of blowouts, the underground blowouts can be the most difficult, dangerous, and destructive situation in well control (Grace, 2003;Barnhill and Adams, 1979). Remedies to UGBs include relief wells (Leraand, et al, 1992 andFlak, et al, 1995), dynamic methods (Gillespie, et al, 1990), injectable pressure-activated sealant technology (Rusch, et al, 2004), and the application of novel lost-circulation material squeeze systems (Sweatman, et al, 1997).…”

Advancements in Dynamic Kill Calculations for Blowout Wells

“….............................. (9). The mixture density in the lower region is the volume-weighted average of the kill-and formation-fluid densities: Pme=PLYu+PgYge .…”