Perforation and Flowback Highlights From the Gorgon Field, Offshore Australia
You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.
The Gorgon liquefied-natural-gas project (Fig. 1 above) is one of the world’s largest natural gas projects and the biggest single resource development in Australia’s history. In 2014, eight Gorgon wells were perforated successfully, intercepting between three and seven commingled zones and gross intervals of up to 500 m per well. This paper contrasts the detailed perforating and flowback plan with the results of the operation where a number of planned, and some unplanned, contingencies were faced.
Perforation Basis of Design
The main element of the Gorgon--project well design that affected perforating was that reservoir sections would be completed with a cemented 70-in. liner in 8¾-in. open hole, with the top perforation approximately 250 m below the liner top. Each well was to be completed in multiple formation zones, with the average gross perforation interval per well (top to bottom shot measured distance) expected to be more than 400 m and an average net perforation interval (sum of perforated zones) of close to 150 m.
Four perforating alternatives were evaluated with respect to their relative operational and subsurface risks:
- Tubing-conveyed perforating (TCP) shoot and drop—guns deployed into the liner on a gun hanger before upper completion is run and then fired and dropped after perforation
- TCP shoot and pull—a separate drillpipe-conveyed perforation trip before completion with use of a sized kill pill after perforation to mitigate losses before the upper completion is run
- Coiled-tubing conveyed
- Wireline conveyed
After analysis, the TCP shoot-and-drop option was eliminated because of the additional time and well-control risk of drilling the required rathole to accommodate dropped guns.
Shoot and pull was not considered a viable option because of a lack of a temperature-rated, post-perforation kill pill.
Coiled tubing was rated feasible but with inherently greater operational risks when performed from a mobile offshore drilling unit.
The process concluded that the wireline-conveyed method was the best solution despite requiring several perforating runs per well. Operational issues associated with this technique were identified, and plans were developed for these to be managed safely.
The gun length per trip was expected to be approximately 30 m, necessitating at least five wireline trips to perforate each well, with each run expected to take between one and two rig shifts.
In the event of failure (e.g., unsuccessful cable development), the selected contingency method for deployment was coiled tubing, which was considered a proven alternative.
Laboratory testing was performed as part of the detailed perforation design for Gorgon. During testing, the merits of dynamic underbalance (DUB) at different applied levels were evaluated against static-overbalanced, static-underbalanced, and balanced perforation cases. The main test result indicated that DUB perforating yields greater productivity and cleaner perforation tunnels.
After testing, the perforating design basis stipulated that DUB perforating guns would be used for all runs, with completion fluid used to provide static-underbalanced conditions for the first perforating run and approximately pressure balanced for the remainder.
The perforating sequence was planned to be top down to ensure that guns were in a liquid column at the time of detonation (a DUB requisite).
Gun-size and individual-perforation-charge selection were validated by nodal analysis.
Two gas gun options were considered, a 4½-in.-equivalent 4.72-in.-outer-diameter (OD) gun and a thick-walled 3½-in. gun. Because 4.72-in. guns are heavier, an average of two additional runs per well are required; however, the same ballistic performance is assumed because the same charge is used as in the 4½-in. gun. The 3½-in. option leads to a small, but relatively insignificant, reduction in gun runs per well because it can be run in 90-ft lengths rather than the 80-ft lengths of the 4½-in. size. This is driven by maximum safe load on the planned high-strength cable. However, the 3½-in. gun comes with a 20% reduction in charge performance. The larger 4½-in. equivalent gun was chosen, with the 4.72-in. type reserved as an option when gas surrounded the guns.
As such, DUB perforating was confirmed and 4½-in. gun size was chosen as the preferred alternative.
Flowback Basis of Design
The well completion fluid was a 9.0-lbm/gal sodium bromide brine. Upon perforation, density differences between the brine and the lower-density reservoir gas were expected to result in a swap of completion brine for gas in the wellbore, compounded as the cushion of highly compressible gas formed below the flowhead and progressively displaced fluid into the open perforations. Earlier studies suggest that this would result in a band of high water saturation accumulating around the wellbore, trapping the aqueous phase in pore throats through capillary forces and leading to an overall reduction in the relative permeability of the gas phase, leading to water blockage. This brine-induced skin would be detrimental to the collection of representative commingled gas samples and meaningful well-productivity assessment, so the minimization of this temporary damage became a priority.
Flowback of the well first would entail an unloading stage, during which completion brine remaining in the wellbore following perforation would be produced straight to flare.
Following the unloading stage, the wells were to be beaned up to a maximum flow rate of 80 MMscf/D, with a set of predetermined stability criteria created to guide the decision on when the flow stream was considered sufficiently representative for sampling purposes.
Perforation. The first well to be perforated was planned as a top-down operation with 4½-in. nominal diameter and 5-shots/ft (spf) guns loaded at 4.5 spf to achieve design DUB. On the fourth and final wireline run, at a depth of approximately 4100-m measured depth from the rotary table, the guns became stuck immediately after perforation initiation. This required activation of the downhole electrical disconnect to recover the wireline tool string alone, without the 34 m of guns and associated wireline tools. A subsequent slickline run indicated that the guns had dropped after release and were in a position that would not restrict the best areas of the perforation interval from contributing to flow. A risk-based decision was made to leave the guns in place rather than attempt to fish. A review of the incident concluded that a combination of perforation-gun and tunnel debris from above and a potential for differential sticking contributed to the issue. Some operational changes were instituted as a result of this finding and were applied to all future runs.
All following wells were perforated with a bottom-up approach to minimize gun-sticking risks, and an unloading flow period was added to the sequence after the initial perforation run to remove brine that otherwise would enter the reservoir. The change in philosophy led to only the initial run being made with DUB guns. Thereafter, because the wells had been switched to gas, approximately on-balance perforation was used. Because these runs were made largely with guns surrounded by gas, a move to the thicker-wall (4.72-in. OD) gas guns was made.
Two separate cable-stranding events were encountered. Recovery of the stranded cable involved identifying the strand-initiation point, working the lower stranded portion back up through wireline pressure-control equipment before dressing off, and then being able to recover spent guns to the surface. As a result of the cable-stranding events, a number of changes were made: larger wireline flow tube size, larger-diameter wireline sheave wheels, reduced running speeds, and a reduction in gun length run to 60 ft.
Despite some of the significant operational issues encountered, all eight wells were perforated successfully with 900 m of loaded gun length and 16,642 shots fired over 69 perforating runs.
Flowbacks. Hydrate formation was an early issue because of high pressure drops of up to 4,000 psi across the well test choke, cool temperatures, and high completion-brine content on startup. When the hydrate would form was unpredictable, and it appeared to form anywhere from the choke through to the flare nozzle. This prompted several contingent reactionary measures, including reductions in flow rate, increases in methanol injection, and alterations in flow-stream direction through the opposing flare boom. Subsequently, all wells were beaned up as quickly as possible to increase the flow-stream temperature and take the gas outside of the hydrate window.
Removing DUB in favor of a bottom-to-top, early-unloading perforation strategy prevented 600–700 bbl of completion brine from entering the reservoir; however, this was countered by the addition of 55 bbl of brine to the wellbore each time spent guns were pulled, to weight up the gas-filled well and create equilibrium across the lubricator valves. With the increase in total perforation runs, this occurred up to 12 times, resulting in up to 600 bbl of completion-brine injection into the reservoir. The knock-on effect was increased water blocking and liquid loading of lower zones, removing them from flow contribution and resulting in additional drawdown.
- Plan to be flexible. Prepare for the unexpected, and be creative in thinking of contingencies.
- Avoid constraining predictions. The creation of meaningful scorecards should make every attempt to reflect the full range of outcomes.
- Use a range of independent methods to tackle nonuniqueness, and acknowledge the limits of data interpretability.
- Be aware of design capability of equipment when operating under new, untested conditions.
- Understand the balance between hydrate management (particularly during cold startup) and the operating requirements of surface equipment and beanup-rate management.
Perforation and Flowback Highlights From the Gorgon Field, Offshore Australia
01 November 2017
After Drop in Activity, Haynesville Shale Sees Resurgence
After a drop in drilling activity in recent years, the Haynesville shale has become a hot area for natural gas production in the US, and companies are looking to bolster their positions in the area.
Hybrid Microseismic and Microdeformation Monitoring of a Coal-Seam-Gas Well
A hybrid downhole microseismic and microdeformation array was deployed to monitor fracture stimulation of a vertical coal-seam-gas (CSG) exploration well in the Gloucester Basin in New South Wales, Australia, to provide more-accurate insight into overall fracture height.
New SCADA Systems Help Operators Fulfill Potential of the Digital Oil Field
Supervisory control and data acquisition systems no longer simply monitor operations and produce large volumes of data in static displays, but now collect production data from all operation data sources and contextualize and present them to workers in real time as meaningful, actionable information.
14 November 2017
13 November 2017
16 November 2017
15 November 2017
14 November 2017