Methodology for long horizontal gravel pack wells in Brazilian offshore deepwater
João Vicente Martins de Magalhães
Agostinho Calderon
André Leibsohn Martins
Petrobras S.A.
Anew methodology for Brazilian offshore deepwater field development provides economic benefits that include heavy-oil fields in ultra-deepwater with 1,000-m to 2,000-m horizontal sections. The method enables openhole gravel packing in long horizontal wells with narrow operational windows to effectively drain the reservoir by maintaining a reasonable productivity index of a heavy oil field.
Many of these heavy-oil reservoirs encompass non-consolidated sandstones, where sand control techniques are required. Sand production causes several surface problems, such as pipe erosion and sedimentation inside the oil/gas/water separator and subsurface equipment.
Many techniques for sand control are available. Economic development of deepwater projects requires that a minimum number of wells be drilled and gain effective reservoir drainage to maintain a high productivity index of the wells.
A viable option for accomplishing this task is to drill long horizontal wells. Openhole gravel-packing of horizontal wells in unconsolidated formations is a very effective way to achieve all of these goals.
The gravel packing technique consists of filling out the annular space between screen and producer formation with sand, ceramic, or other solid particles with selected grain diameter. The idea is to create a second porous medium with a pore throat diameter smaller than the formation grain diameter and, in this case, when the well is in production time, oil would easily flow through the gravel pack while formation particles would not.
Gravel pack placement is a risky job. Usually, in ultra-deepwater, the layer of sediments is very thin, resulting in a low formation fracture resistance. Pumping gravel in long-horizontal-section wells generates high dynamic pressures in the open hole. These aspects were the main motivators for a software development that calculates the pressure drops during gravel pack placement and optimizes operational pump rate.
Flow rate limits
To achieve a successful gravel pack operation, various hydraulic limits have to be acknowledged. Dynamic pressures during the operation should be maintained between a window formed by the pore pressure and the fracture pressure. If the wellbore pressure, at any time, is below the pore pressure, there will be influx of formation fluid to the well. On the other hand, if the wellbore pressure is greater than the formation fracture pressure, there will be influx of completion fluid to the formation, possibly generating damage.
Another important issue is to guarantee that the operation will be run at a minimum flow rate that avoids premature screen out of the rat-hole. Since a larger diameter openhole section is exposed, if the flow rate is too low, alpha waves formed may be high enough to block sand passage to the openhole, generating immediately a beta wave in the rat hole. The consequence is that the pressure at the last casing shoe will immediately increase and the operation will have to be aborted without packing the openhole.
With the input data, the computer program can predict alpha wave heights in the openhole and rat-hole sections, besides pressure propagation during injection/alpha wave/beta wave stages with displacement time.
These results generate outputs in two different ways:
• Defining the operational window based on the minimum flow rate required to avoid premature screen out in the rat hole and the maximum flow rate that does not lead to formation fracture. Normally, the fluid density is designed to generate wellbore pressures in static conditions higher than pore pressures
• Predicting the pressure propagation with placement time for a fixed flow rate.
Case study definition
To support a case study, a typical well design for deepwater was simulated that considered critical conditions such as: 2,000 m water depth, 2,000 m horizontal openhole length and low frac gradient. The input data for simulation was:
• Open hole diameter: 8 1⁄2-in.
• Rat hole diameter: 12 1⁄4-in.
• Water depth: 2,000 m
•Las casing shoe: 3,175 m (MD); OD = 9 5⁄8-in.; weight = 47 lb/ft
• Open hole length: 2,000 m
• Reservoir TVD: 2,929.72 m or 9,611.9 ft
• Screen: OD = 6.13-in.; ID = 4.89-in.
• Screen position: centralized
• Column: OD = 5-in.; weight = 19.5 lb/ft
• Wash pipe: OD = 4-in.; ID = 3.48-in.
• Kick-off point: 2500 m
• Build-up rate: 4°/100 ft
• Open hole inclination: 90°
• Fluid density: 9.5 ppg
• Gravel concentration: 1 ppg
• Gravel type: sand 20/40
• Fracture gradient: 0.56 psi/ft
• Pore pressure ECD: 8.60 ppg
• Fracture pressure ECD: 10.77 ppg
• Open BOP configuration - reduces friction losses since return flow happens through the riser, where negligible friction occurs due to its large diameter.
The software simulation outputs for this case study indicated that the operation seems to be unfeasible. The maximum pump rate to avoid formation fracture, at the end of beta wave propagation, is 3.0 bpm, while the minimum pump rate, to prevent rat hole premature screen out, is 4.1 bpm. These results indicate that there is no operational window for the job unless special design strategies are considered.
Operational strategies
Several strategies were formulated to enlarge or create operational windows in critical operations. The strategies were simulated separately to compare the individual impact on the operational pressure. Each technique represented different levels of risk, costs, and operational complexity.
Normally, the fluid density is designed so that the hydrostatic pressure generated is higher than the pore pressure and lower than frac pressure. Once the operational window gets narrower in the well, dynamic pressures will overcome frac pressures. One alternative is to reduce the fluid weight to levels that guarantee the total gravel pack displacement in the extended well.
In the case study, the fluid weight was reduced until a value that guarantees 200 psi of overbalance plus pore pressure resulted in 8.60 ppg. However, the calculation determined that the fluid weight should be 9.0 ppg. Operations with fluid weights ranging from 9.5 to 9.0 ppg were also simulated that indicated, even for the lower fluid weight, there was no operational window. Although the reduction of fluid weight by itself does not make the operation feasible, the strategy may be applicable when combined with other alternatives.
Currently, a large range of solids are available as alternatives for annular packing. Various kinds of solids with different shapes and specific gravities will provide varying alpha wave heights and, consequently, different operational windows.
Significant improvement was achieved with the use of lightweight proppants. Due to its smaller tendency to sedimentation, premature screen-out of the rat hole is avoided at low flow rates.
A second light proppant simulation was performed after lessening the security criteria of 200 psi overbalance plus pore pressure. Using an 8.6-ppg fluid, a 2.5-bpm operational window was obtained.
Zero rat-hole
The rat-hole is a critical issue in the operational window definition. The casing design community is attempting to make possible a zero rat-hole well or, when this is not possible, to put down devices that will bypass the rat-hole. In this case, the lower operational limit (minimum flow rate) would be reduced, and an extended well operation could be run with lower flow rates.
An additional issue to be addressed is the tendency of alpha wave deposition inside the work string when pumping at low flow rates.
Beta wave placement pressures can be reduced by the use of flow divergence valves. In this concept, when the pressure reaches a pre-defined value, flow is diverted directly from the openhole to the interior of the washpipe, reducing the fluid path dramatically and, consequently, minimizing friction losses. In front of low frac formations, where the dynamic pressures reach the fracture pressure, these devices can be an interesting alternative to enable beta wave placement and to enlarge operational window. Operating with narrow windows is risky, due to uncertainties in parameters, in addition to fluctuations that naturally occur while pumping the slurry. Combining three flow divergence valves with light proppants, will increase the operational window to 4.0 bpm.
Drag reducers are polymeric additives added to a fluid to reduce friction losses during pumping. Turbulence suppression and decrease of kinetic energy transport are possible mechanisms for the phenomenon. Efficient drag reducers can minimize friction losses up to 80% in clear solutions and up to 20% in solid suspensions.
During the placement of the gravel pack, friction losses are generated both by the solids suspension (at the downward flow through the string and at the annular flow through the open hole/screen section) and by clear solution (at the annular flow through the screen/wash pipe section and at all the return flow through the wash pipe, casing annulus, and kill/choke lines or riser annulus in case of open BOP configuration).
To exploit heavy oil reservoirs, an alternative is to drill wells with 9 1⁄2-in. in the openhole section. The larger diameter decreases friction losses during the gravel pack placement operation. Operational windows are much wider when compared with similar 8 1⁄2-in. wells simulation. Even in this case, different strategies should be combined to provide comfortable operational windows.
Large alpha wave
During horizontal openhole gravel packing placement, the beta wave phase is critical, because it generates abrupt pressure increases due to the divergence of annular flow from the open hole/screen section to the narrow annular screen/wash pipe section. A last alternative to make possible openhole gravel packing at extended horizontal wells would be to design a large alpha wave that would cover the screen without significant pressure increase. The beta wave stage would be initiated and performed until frac pressure limits allows.
By this point, the operation would be interrupted and when the pressure is alleviated, the non consolidated formation would collapse on the top of the alpha wave at the regions which the beta wave could not be placed. Theoretically, the formation material would not reach the screen since the operation was designed for total screen coverage by the alpha wave.
Certainly, this is a risky strategy, but it can allow annular packing in extreme cases, where all the other alternatives are not applicable, available or would not create an operational window.
The strategies consist of effective alternatives to extent hydraulic limits and achieve success in critical operations. Some of the alternatives are associated with reasonable risk and should be considered only in situations where no other alternative was successful. The proposed methodology represents a powerful tool for complex scenario for gravel pack design.•
Acknowledgements
The authors would like to thank Petrobras S.A. for permission to publish this paper.
Editor’s Note: This is a summary of a paper presented at the Deep Offshore Technology Conference & Exhibition held Nov. 8-10, 2005 in Vitoria, Brazil.
