GEORGE GILLESPIE, CHRISTOPHER HALL, STEPHEN McNAMEE and JOHN SLADIC, Weatherford International; CHARLES YEH, Exxon Mobil Upstream Research Company
 |
| Lateral compression tests simulated the effect of downhole compaction by the reservoir formation. Test results suggest that the MazeFlo screen is as strong as a conventional wire-wrapped screen, and the compaction would not cause additional damage to the secondary screen. |
|
A significant improvement in screen reliability has been shown in laboratory and field tests of a unique, self-mitigating sand screen technology. Developed through a joint project between Exxon Mobil and Weatherford, the new technology enhances performance by compartmentalizing sand ingress in a maze of baffles between two screen layers. Weatherford has a license from Exxon Mobil on this patented sand control technology.
Rigorous testing has verified that the mechanical performance of a 2 3/8-in. MazeFlo screen is not compromised by the self-mitigating design compared to a conventional screen. Field-testing confirmed the commercial viability of the screen as well as the self-mitigating capability at the designed rate capacity.
The 2 3/8-in. MazeFlo screen is now being deployed in the field and further design efforts are focused on development of a 5½-in. screen for primary completions that can also be combined with other emerging technologies.
CRITICAL NEED
High-rate production, combined with high pressures and temperatures, deep water, remote locations, long intervals and multiple zones makes sand control reliability critical to ensuring completion integrity. However, despite much advancement, screen failure is still a risk for maintaining production levels. Failure prediction is difficult because plugging and erosion failures are partly driven by downhole uncertainty.
Conventional efforts to mitigate failure include preventive (flux limitations) and reactive (choking) methods. While these methods can reduce the risk of downhole damage under complex producing environments, they are unable to eliminate all the risk.
Exxon Mobil’s patented MazeFlo concept (Fig. 1) is a new approach to mitigation-based sand control on three key points: adapting to downhole uncertainties, self-mitigation of local screen damage and greater overall reliability. The screen design features a series of flow compartments along a selectively perforated base pipe. Each compartment has a primary screen, outer housing, flow baffles and a secondary screen.
 |
| Fig. 1. Screen reliability is enhanced by the novel use of a series of flow compartments along a selectively perforated base pipe. |
|
Wellbore fluid flows into the primary screen and is redistributed by flow baffles. The fluid continues to the secondary screen and then into the perforated base pipe, where it commingles with production from the other compartments. Should an erosion “hot spot” occur in the primary screen, sand then flows into the annular compartment and accumulates on the secondary screen. This accumulation creates a flow resistance in the problematic compartment, diverting fluid to the surrounding, undamaged screen compartments. This self-choking action occurs automatically and only at the sand breakthrough location, without surveillance or control. Conventional reactive methods would choke the entire producing interval.
PROTOTYPE DEVELOPMENT
The prototype screen uses a 2 3/8-in. base pipe with direct wire-wrap primary and secondary screens. Each screen joint has three, 10-ft maze compartments in series. Blank spaces adjacent to the compartments support conventional elevator handling and make-up operations.
Physical testing was conducted to qualify hydraulic performance (friction flow, erosion and pack-off tests) and mechanical integrity (collapse, burst, bending, torsion, tension, compression, push-off and load bearing). The sizing of the housing/secondary screen annulus and the baffle region was key to achieving a balanced and uniform flow distribution. Friction loss, erosion resistance, and sand pack-off within the compartment were qualified by the hydraulic flow tests, with extensive analyses and erosion modeling using computational fluid dynamics (CFD).
HYDRAULIC PERFORMANCE
Hydraulic flow tests were conducted to qualify friction loss, erosion resistance and sand pack-off within the compartment. The results verified that the new screen design does not induce significant friction loss in redistributing flow. This low compartment friction loss was estimated by CFD modeling and verified by flow testing.
The friction test procedure pumped water through the annulus between the primary screen and the non-perforated base pipe, into the baffle section, through the secondary screen and into the perforated base pipe. The loss across the baffle section was less than 3.5 psi over a wide range of flowrates up to 1,200 bpd through a single-maze compartment. In the baffle section, the friction loss was about 1 psi, assuming a 720-bpd rate through a single 2 3/8-in compartment across a 10-ft interval.
Erosion testing was performed to evaluate plugging potential and erosion on the baffle section and the welded areas during a primary screen failure. To produce failure conditions, test sand was loaded into water and injected through an entry port at a rate of 1,440 bpd.
A high-rate sand injection port was used to impose a worst-case erosive condition on the maze compartment. The port covered about 15% of the screen circumference to produce a high-rate, concentrated sand inflow. No visual erosion was observed in the compartment and its welded areas passed the post-test inspection. The test results verified that the compartment design has minimum erosion or plugging potential.
Pack-off tests (Fig. 2) helped verify sand pack-off in the compartment in the event of primary screen damage. A transparent housing was installed to view the pack-off action. Test sand was loaded in water and injected at 720 bpd into the compartment. The test was stopped after about 2 min. when the pack-off reduced initial flow by more than 97%. The time was significantly less than required for passed sand to start eroding the screen. Under downhole conditions, the wellbore fluid would divert to adjacent compartments.
 |
| Fig. 2. Extensive testing included analysis of compartment pack off when the primary screen is damaged. Chart shows the ratio of flow capacity during sand injection. |
|
MECHANICAL TESTS
Mechanical qualification efforts tested prototype screens manufactured with L80 base pipe and a 9-gauge (0.009-in.) slot size. The prototypes were manufactured to Q1 quality grade, ISO Standard 17824. The objective of a rigorous testing program was to verify that the mechanical performance of the screen was at least equivalent to a conventional, direct-wrap screen or was limited by the base pipe similar to a conventional screen. Results showed that the new screen is as strong as a conventional wire-wrapped screen and is similarly limited by the base pipe.
Burst and collapse tests were performed by pumping a fluid loss control pill to seal the slots and apply hydraulic pressure either internal or external to the screen. Three test assemblies were fabricated: a conventional screen, MazeFlo primary screen plus housing (no secondary screen) and MazeFlo secondary screen plus housing (no primary screen).
For an 8-ft conventional wire-wrapped screen, the peak burst pressure was 3,865 psi. In burst tests of the MazeFlo primary and secondary screens, the peak pressure was at approximately 4,200 psi, which is comparable to or higher than conventional screens of the same configuration.
The peak pressure in the collapse tests for an 8-ft conventional wire-wrapped screen was 5,223 psi. Collapse testing ended with hot spot erosion that followed multiple attempts to recover the applied pressure after loss of sand control. No physical collapse or significant geometric change was observed, even though the screen had eroded.
The MazeFlo secondary screen had a higher peak collapse pressure of 5,596 psi, possibly due to its shorter length. Testing of the primary screen and housing resulted in collapse of the housing prior to the primary screen at about 3,500 psi. This rating can be improved by increasing the wall thickness. Push-off testing determined the integrity of the components relative to the integrity of the base pipe when pulling or pushing to free a stuck screen. The components linked to the self-adapting capability (secondary screen and housing) were not a limitation to overall mechanical performance. Results indicated that the secondary screen would maintain integrity even after failure of the primary screen’s end ring to base pipe weld at 93,000 lb.
Tensile testing was performed to determine screen limits when running in or pulling out of the wellbore. A screen compartment was positioned horizontally in the load frame and loaded under axial tension. Displacement transducers were installed to monitor the movement of various components during loading. The test was stopped at several predetermined loading points for inspection.
Inspection indicated that all welds and screen slots associated with the secondary screen and the housing met criteria. The tensile resistance of the new screen design is equivalent to a conventional screen with the yield failure point limited by the base pipe. The screen's self-adapting capacity also offers additional protection when the primary screen pulls out from the end ring during plastic elongation past the yield point. Bend testing was done to evaluate the flexural strength when the screen is run through a dogleg restriction in the wellbore. The results showed no loss of sand control integrity in doglegs exceeding 20°/100 ft, which is more than most hole geometries.
Torsion testing determined the maximum torque to apply should pipe rotation become necessary to free a stuck screen. A maximum torsional load of 6,000 ft-lb was applied to the screen. After the test, the secondary screen and welds in the maze compartment were inspected and all components were within the critical acceptance criteria.
Lateral compression tests simulated the effect of downhole compaction by the reservoir formation. The objective was to determine if the housing over the secondary screen introduced any additional loading points to the secondary screen that would be detrimental to the self-mitigating functionality. Testing validated that the MazeFlo screen is as strong as a conventional wire-wrapped screen and that compaction of the housing would not cause additional damage to the secondary screen.
FIELD TEST
Following the successful prototype development and qualification, a 2 3/8-in screen was subjected to a field trial to demonstrate the screen installation, downhole self-mitigating capability, and sustained production. The field test was an extension of the product qualification process in a practical time frame. The test was conducted in a sand-prone well with high local flux. It had multiple objectives, including successful installation, early sand-free production at the target rate, demonstration of self-mitigation via a pre-damaged primary screen, demonstration of self-adapting capability via forced erosion of the primary screen, successful screen retrieval and inspection and determining the rate capacity for self-mitigation.
The field test was designed to subject the screen compartments to production rates that far exceed normal operation. This was done in two ways: by minimizing the number of compartments to concentrate the flow and limit the flow paths; and by producing the well at higher than recommended rates given the minimal number of installed compartments.
The subject well was originally completed and perforated through 5½-in. casing over a 12-ft interval at 5,757-ft MD. It was producing with an electrical submersible pump (ESP) drive, but had sanded up and fill was tagged above the perforations.
After pulling the ESP and cleaning sand out of the wellbore, the well was recompleted with MazeFlo screens and the ESP was reinstalled. Three 20-ft screen joints were installed in the well. Each contained one maze compartment with a 132-in. primary screen and a 30-in. secondary screen. A 9-gauge (0.009-in. or 229 micron) screen slot size was selected based on sand grain size. The primary screen was positioned at the box end of the joint and the secondary screen and housing on the pin end.
The primary screen on the bottom joint was intentionally damaged at the factory to test the self-mitigating capability. The primary screen on the middle joint was located across from the perforated interval, and the top joint was located above the perforations. The screens were installed in a vertical well, which produced at 2,500 to 3,000 bfpd, sand-free with a high water cut for three months. At that point, a small amount of fine sand was produced, and a few days later, the ESP sanded up. The screens were successfully retrieved and inspected. Other than the factory-created test damage on the bottom joint there was no damage to the primary screen. Removal of the housing (Fig. 3) showed three quarters of the secondary screen packed with sand. No erosion was observed. All slots were within the tolerance, and welds passed LDP tests.
 |
| Fig. 3. Sand packing of the secondary screen in the bottom joint shows successful trapping of sand ingress at an intentionally created “damage” location. |
|
The middle joint primary screen was eroded about a foot from the housing, which was probably the result of a concentrated sand jet from the perforations. The secondary screen was, again, largely packed with sand. An eroded hot spot occurred near the leading edge of the secondary screen, but no other erosion was found. All the welds passed the LDP tests.
The top joint’s primary screen and base pipe were severely eroded and the lower quarter of the secondary screen was covered by sand. The secondary screen was eroded at several spots and there was some erosion through the base pipe. However, all the welds passed the LDP test and no erosion occurred in the housing or baffle area.
Our analysis concluded that the intentionally damaged bottom compartment packed first while the production was distributed equally through all three joints via the open annulus. Approximately 800 to 900 bfpd of sand-laden fluid flowed through the bottom joint and packed the compartment without eroding the secondary screen.
Once the bottom compartment was packed, the flow was redirected through the two remaining compartments at a higher rate in an effort to induce erosion on the middle compartment’s primary screen. The location and the size of the eroded area indicated that the sand ingress and packing occurred at a slower rate than with the bottom joint. The flow-induced erosion was considerably smaller compared to the factory-induced test damage.
As the bottom compartment was being packed, production gradually shifted to the middle and top joints, forcing a top rate of 1,500 bfpd of sand-laden fluid through the middle and top compartments. This flow, combined with the primary screen location directly across from the perforations, resulted in a hot spot on the primary screen and progressive packing of the middle compartment. The extreme production rate combined with sand caused one eroded hot spot on the secondary screen, which was expected, given the severe nature of this test.
After the middle compartment was packed, all 3,000 bfpd of production was directed through the only remaining open flow path—the top joint. This extreme rate found a path through the primary screen next to the housing and into the compartment, causing erosion on the primary screen, the secondary screen and the base pipe. The concentrated erosional conditions allowed sand to enter and fill the base pipe, and ultimately the sand reached the surface and plugged the ESP in the final days of the test. This was an expected outcome given the limited compartments and high production rates designed to provide the team with a definitive signal to conclude the test and pull the screens.
The field test results describe a self-mitigating or pack-off capacity per maze compartment of about 800 bfpd, which is significantly higher than typical production for a 23/8-in. screen completion. This design is primarily used for openhole completions, remedial for perforated liners, failed SAS or openhole gravel pack completions, and the new 5½-in. design is best suited for initial completions.
All six of the field test objectives were achieved. Had the objective been to actually maintain sand-free production, more compartments would have been installed to increase available flow paths and reduce the flowrate per compartment. In addition, compartment length can be varied by adjusting the primary screen length while keeping the secondary screen unchanged. The commercial screen design has three maze compartments per joint of pipe. In higher rate wells or more heterogeneous reservoirs, inflow control devices (ICD) can also be installed to further control the production rate per compartment.
NEW FIELD OPTIONS
Both laboratory and field qualification programs have established the readiness of the 23/8-in. MazeFlo screen for field applications. The knowledge gained during this extensive development and qualification process ensures proper design and capacity to meet field objectives. By self-adapting to the downhole uncertainty and self-mitigating screen damage, the new screen technology enhances reliability in standalone screen and gravel pack completions. Greater reliability benefits well production sustainability and results in more completion design flexibility in balancing productivity, performance, cost and operational complexity.
Dedicated collaboration between the operator and service provider enabled the development of the technology from engineering design, prototype fabrication, qualification testing, and production manufacturing through to field application. The development of the 2⅜-in. screen has demonstrated the successful collaboration between an operator and a service provider. This teamwork has delivered an innovative technology to address the complex challenge of sand screen reliability. 
LITERATURE CITED
1.Yeh, C. S., et al., ”Enhancing sand screen reliability: An innovative, adaptive approach,” SPE paper 134492, presented at the SPE Annual Technical Conference and Exhibition, Florence, Italy, Sept. 19-22, 2010.
2.Yeh, C. S., et al., “Unlocking the Limits of Sand Screen Reliability with an Innovative and Self-Adapting Technology,” IPTC paper 14623, presented at the International Petroleum Technology Conference, Bangkok, Thailand, Feb. 7-9, 2012.
ACKNOWLEDGMENT
MazeFlo is a trademark of Exxon Mobil Corporation.
|
