Long, horizontal gravel packs are viable completions that have been placed successfully in more than 80 wells. Extensive field-scale testing has significantly aided the development of field procedures and operating guidelines. Software has also been developed to assist with horizontal gravel-pack design. To be performed properly, these completions require a systems approach for their implementation because drilling and displacing the completion interval, maintaining hole stability, selecting and running equipment, and maintaining returns during the gravel-packing operation must be integrated into the completion strategy. Field experience suggests that, in some cases, gravel packs maintain productivity better than prepacked screens or slotted liners.
Gravel packing is a well-completion technique used to exclude formation sand from produced reservoir fluids and extend completion longevity. It is performed by placing a gravel-retention device (such as a wire-wrapped screen) in a well opposite the completion interval, and subsequently circulating gravel around the gravel-retention device to create a permeable filter. Gravel packing has been applied in vertical and deviated wells that were completed either cased or openhole. Transport fluids used to gravel pack wells have been either polymer-viscosified brines or unviscosified brines. Of the two choices, unviscosified brines are usually preferred because they exhibit lower-porosity, void-free packs.
Several authors have investigated the factors affecting gravel placement, productivity, and completion longevity.1–17Fig. 1 illustrates an example of the gravel placement at well deviations of 0 and 45° using water as the transport fluid.16 Note that the gravel is initially placed at the bottom of the well and packs sequentially upwards. At well deviations in excess of 60°, the angle of repose of gravel is exceeded and the gravel initially settles on the low side of the well and remains at rest, unless sufficient flow is available to move it forward. The 60° well deviation represents a transition in the placement of gravel, because at lower well deviations the gravel will settle to the bottom of the well.
Placing gravel at the bottom of a highly deviated well requires higher pump rates and a slightly different completion geometry than in vertical wells. A large-diameter wash pipe can be used inside the screen to enhance placement.2 It forces a larger fraction of the fluid to flow along the outside of the screen, rather than being diverted into the screen/wash-pipe annulus. The higher screen-casing annulus flow rate assists in transporting the gravel and prevents a premature stall of the gravel pack before it reaches the bottom of the completion. When the proper screen and wash-pipe size are used, the gravel-deposition process consists of a primary or "alpha" wave that packs from the top to the bottom of the completion, leaving an open void over the gravel dune. Upon reaching the bottom of the completion, a subsequent secondary or "beta" wave deposition proceeds in the opposite direction of the alpha wave (towards the top of the well) until the interval is completely packed. The packing process in a highly deviated well is portrayed in Fig. 2.16 This technology has been used to successfully gravel pack many conventional wells with deviations as high as 80° or more and completion lengths of several hundred feet.
Until recently, gravel packing long, horizontal wells has not been used as a completion technique, apparently because the technology was not thought to be in hand. Instead, prepacked screens and slotted liners had been used as stand-alone sand-control equipment, even though their performance in conventional wells had been disappointing. Their improved performance in horizontal wells is probably attributed to much lower flow rates per foot of completion interval than when used in vertical wells. In horizontal service, the initial productivities of screens and slotted liners usually have been good, but their tendency is to plug and restrict productivity, particularly when completed in dirty sands, when high water or gas/oil ratios occur, or when placed in viscous-oil service. Not surprisingly, screens and slotted liners seem to perform best in clean, high-permeability sands. Hence, how well screens and slotted liners perform tends to be site-specific.
Gravel packing horizontal wells has been successfully demonstrated in field-scale studies and actual completions. Gravel has been placed in over 80 wells, with intervals ranging from 600 to 3,300 ft. The test data and field experience suggest that packing even longer intervals is possible. The following discussion reviews testing and field experience gained with gravel packing long horizontal wells. It also presents guidelines for completion operations.
Most field-scale testing for simulating gravel packs has been performed with clear plastic models. Until recently, the gravel-pack-model lengths were less than 100 ft, and most were about 25 ft long. The implications of testing in these models were that they were too short to provide meaningful results for long, horizontal wells. While numerical simulation was also an alternative, designing completions using unverified simulators was believed to be risky and unreliable. For these reasons, a field-scale model was designed and constructed to simulate gravel packing horizontal wells.
The horizontal model is 1,500-ft long with a 4 1/2-in. outside diameter (OD) [4-in. inside diameter (ID)]. It is equipped with a centralized 2 1/16-in. screen (0.006-in. slot openings). The wash-pipe diameter is 1.315 in. The wash-pipe OD to screen ID ratio is 0.75. The model is equipped with clear, high-strength plastic windows capable of operating at pressures of 1,000 psi to allow the visual observation of the packing process at six locations along its length. Sliding sleeves are also located along the model to allow the visual observation of gravel deposition in the model at that location after pumping has ceased. Fluid loss is simulated with 400 perforation tubes that are l-ft long, 1/2-in. diameter pipe filled with resin-coated 40/60 U.S. mesh gravel. Five pressure sensors are installed at strategic locations. Inlet and exit flowmeters measure entrance and return flow. Pumping is performed with a field unit. Data acquisition consists of recording rates, pressures, and gravel-mix ratio as a function of time. Fig. 3 illustrates a schematic of the model while Figs. 4 and 5 show the windows and the perforation arrangement.
On the basis of previous experience, water was chosen as the carrier because viscous fluids have a strong tendency to bridge and dehydrate the gravel prematurely in long, highly deviated wells.16