Horizontal completions-Conclusion

Gravel packing prevents productivity decline

W.L. Penberthy Jr., Kenneth L. Bickham, H.T. Nguyen
Baker Oil Tools
Houston

Gravel packing horizontal wells has the advantage of creating superior productivity without the productivity decline observed with stand-alone screens.

This conclusion of a three-part series, which began in OGJ July 21, 1997, shows how extensive field-scale testing has aided the development of procedures and operating guidelines for gravel packing horizontals.

Experience has been that open hole gravel packs require a systems or process approach to achieve a successful completion. The delivery of a clean, stable, undamaged well bore to the gravel-pack phase aids significantly in achieving a high-productivity completion.

To gravel pack a long horizontal well requires sound technology that can be applied in the field using simple procedures. Field-scale research has developed horizontal gravel-pack technology that has subsequently been reduced to procedures that have been applied to gravel packing about 100 horizontal wells, the longest of which is over 3,300 ft in length.

Gravel packs

Gravel packing offers another option for completing a horizontal well where sand production is a problem.

However, the general perception has been that technology was not available for gravel packing long, horizontal wells, and that other alternatives, such as stand-alone screens must suffice. This is contrary to the fact that the performance of stand-alone screens has been unacceptable in conventional wells. Moreover, data from stand-alone screen completions in horizontal wells indicate that over 25% can be classified as mechanical failures because they no longer control sand production or produce at commercial rates.

However, one of the most disturbing facts as previously mentioned is that the average pressure drop across the remaining stand-alone screen completions is in excess of 700 psi. The obvious implication is that the screens are plugging because they function as a downhole filter rather than a gravel-retention device. This is consistent with the data discussed in the second part of this series (OGJ, Aug. 11, p. 63).

Integrated completions

While extensive literature is available on gravel packing vertical and deviated wells,1-18 horizontal gravel packs are less publicized because it is new technology. To effectively gravel pack an open hole horizontal well involves a drilling/completion process that is significantly different than cased-hole completions.

It requires substantially more than the actual gravel packing. The process entails drilling the open hole, displacing the drill-in fluid to brine using the methodology described in the first part of this series, and finally, gravel packing the completion.

Probably the failure to drill the completion interval with a compatible fluid and to displace the horizontal open hole with brine is the most disregarded portion of the process. One must keep in mind that the key to gravel packing is a clean, stable, undamaged well bore. Modifying proven procedures usually translates into less than the desired well performance.

To date, about 100 open hole horizontal wells with lengths ranging from 600 to over 3,300 ft have been gravel packed. Gravel packing long horizontal wells is actually an extension of existing gravel pack technology that requires effectively cleaning and displacing the open hole, running properly sized equipment, and maintaining return flow to the surface through a wash pipe.

Most horizontal gravel packs have been in open hole completions that were drilled from the casing shoe to the toe with special nondamaging, calcium carbonate drill-in fluids that form a thin, 1-mm thick, filter cake. Provided that a clean, stable, undamaged borehole was delivered to the completion phase, all wells have been successfully completed.

The subsequent productivity of horizontal gravel packs has been substantially superior to stand-alone screens as follows:

• Stand-alone screens-700 psi pressure drop, rates to 1,500 bo/d
• Horizontal gravel packs-100 psi pressure drop, rates to 6,000 bo/d
• Open hole gravel packs-less than 10 psi pressure drop, rates to 7,000 bo/d/ft.

The gravel packs have not experienced productivity declines when performed properly.

Field-scale tests

To determine the feasibility of gravel packing a long, horizontal well, one needs to assess the completion equipment design, pumping schedules, and other related procedures with scaled physical models.

Well deviations of up to 60° tend to initially assist in transporting the gravel to the bottom of the completion interval (Fig. 1 [20,604 bytes]). However, at deviations exceeding 60°, the angle of repose of the gravel is exceeded, and dimensional changes must be made to the gravel-pack equipment. This means that higher pump rates are required to completely gravel pack the entire interval.

The main requirement is that the wash pipe OD to screen ID ratio must be at least 0.75 and returns through the wash pipe must be sufficient to transport the gravel to the toe of the well.

The gravel placement at deviations exceeding 60° is initiated at the top of the completion interval rather than at the bottom of the well when deviations are lower because the angle of repose of the gravel has been exceeded.

Fig. 2 [21,754 bytes] illustrates the gravel placement sequence from initial deposition extending downwards from the top of the completion interval until the gravel dune, commonly referred to as the alpha wave reaching the bottom of the well (shown by the Time Sequence 1-10).

At that point, secondary placement or beta-wave deposition packs the volume above the alpha wave as shown by Time Sequence 11-15. However, the alpha wave will prematurely stall if the gravel concentration is too high, the flow rate is too low, or the wash pipe permits excessive flow in the annulus between it and the screen.

Increasing the diameter ratio to 0.75 and maintaining a return flow superficial velocity of 1 fps promote the stable alpha-beta wave packing sequence (Fig. 2).

Studies in a 7-in. OD by 25-ft long scaled gravel-pack simulator have confirmed the findings shown in Figs. 1 and 2. However, because the model was short, there was concern that horizontal gravel-pack test results would not be representative for actual conditions because tests could be dominated by end effects. Consequently, a longer field-scale model was designed and constructed.

The model was 1,500-ft long, with a 41/2-in. diameter, and contained wire-wrapped screen and wash pipe (Fig. 3 [16,247 bytes]). Fluid loss was simulated by using 1-ft long pipe filled with resin-coated gravel. The difference in the flow into the model and the returns through the wash pipe was the fluid loss to the formation.

The model was equipped with high-strength plastic windows that allowed the viewing of the gravel packing process as it progressed down the model. Fig. 4 [16,363 bytes] shows the alpha wave traversing a window.

Fig. 5 [16,111 bytes] locates typical alpha and beta waves as a function of time. It demonstrates that the entire 1,500-ft model was packed with gravel.

Testing clearly revealed that the alpha-wave height was not constant with pack length as had been implied from studies conducted in 25-ft models. Instead, the alpha-wave height was inclined upwards from the heel to the toe of the model (Fig. 6 [13,860 bytes]).

The reason for the inclination is that fluid loss reduces the annular flow velocity and increases the gravel concentration, thereby reducing the gravel-transport efficiency. The consequence was an increase in the alpha-wave dune height with length.

If the top of the borehole interferes with deposition over the top of the alpha wave, deposition stalls and beta-wave deposition begins at the stall location. To avoid a premature stall, the superficial annular velocity must be maintained above a minimum value, 1 fps, based on return flow through the wash pipe.

The superficial velocity is defined as the ratio of the return flow rate through the wash pipe to the annular area between the screen and the well bore.

Provided that the gravel-pack design is sized properly and a superficial velocity of 1 fps is maintained, gravel packing a long horizontal open hole can be a routine procedure. However, for open hole completions, a clean, stable well bore is an additional requirement for a quality gravel pack that avoids being contaminated with formation material.

The drill-in and displacement procedures discussed in Part 1 of this series have worked well in achieving a clean, stable well bore provided that active shales are not in the flow path.

Fig. 7 [25,746 bytes] illustrates the recorded pressure, rates, and gravel concentration as a function of time for a field-scale test in the 1,500-ft model. For this test, the pump rate was 1.5 bbl/min with 0.75 bbl/min return flow through the wash pipe. The entire model was easily gravel packed.

After pumping about 2 hr, pressure-vs.-time had a distinct increase in slope. The slope change reflects the end of alpha wave deposition and the initiation of beta-wave deposition.

Fig. 8 [17,028 bytes] illustrates data acquired from an actual completion, a 2,500-ft horizontal gravel pack. Observe that for this well, the pump rate was about 5 bbl/min and the return rate was 4 bbl/min. This is typical for most horizontal gravel packs performed to date.

Also note the slope change of the pump pressure-vs.-time at about 61/2 hr into the gravel pack. This also signals the initiation of beta-wave deposition.

The similarities in these data and those shown in Fig. 7 are not unique to these two examples and are a routine pressure-vs.-time signature for horizontal gravel packs.

Field results

Table 1 [21,525 bytes] and Table 2 [14,498 bytes] tabulate the completion and job execution results of selected horizontal gravel packs. All horizontal gravel packed wells were completed open hole and most used 40/60 U.S. mesh gravel and prepacked screens. The deepest well was 8,500 ft. The longest pack was 3,300 ft, a record length to date. Most wells had horizontal lengths of 1,500-2,000 ft. Well bore diameters ranged from 4.75 to 8.5 in.

The gravel packed wells have not experienced the productivity declines observed with stand-alone screens, provided that the completion process described previously was followed. For example, the wells in the 10-well project described in Tables 1 and 2 have significantly better productivity indices (PI) than wells in the same projects that were completed with stand-alone prepacked screens.

Typical gravel concentrations pumped were about 1 ppg; however, pack execution times were reasonably short except for large diameter holes. Typical gravel-pack times are in the 4-6 hr range.

The second project has 50 wells with horizontal gravel packs, and an additional 20 wells remain to be gravel packed. Ironically, the initial completions were stand-alone slotted liners. Their initial flow rates were 3,000-5,000 b/d, but sand control was excessive and the wells plugged with the onset of water production. All stand-alone slotted liners failed.

By gravel packing the wells, productivity has been maintained even after water breakthrough.

Note that the third project represents the average of 11 wells in a particular field that were gravel packed. Their productivity has been excellent; however, about half way through the project it was decided to run a stand-alone 40/60 U.S. mesh prepacked screen to determine if gravel packing was actually needed.

The stand-alone pre packed screen completion experienced a significant productivity decline from the onset. After several months, the screen was removed and the well was gravel packed. The ensuing productivity was superior to the prepacked screen completion, consistent with the other gravel packed wells, and did not experience the productivity decline of the stand-alone prepacked screen completions.

This particular project was subsequently acquired by another operator who drilled and completed three additional wells with stand-alone, proprietary, multilayer sintered metal screens on the assumption that they would not plug like the prepacked screens. The initial productivity from these completions could not be sustained and the well performance has been disappointing and has declined with time. As a consequence, the operator removed the screens and gravel packed these wells.

Several horizontal gravel packs are currently in the execution or planning stages. Some planned wells have completion lengths up to 4,000 ft and others will be performed subsea.

References

1. Maly, G.P., Robinson, J.P., and Laurie, A.M., "New Gravel-Pack Tool for Improving Pack Placement," JPT, January 1974, pp. 19-24.
2. Gruesbeck, C., Salathiel, W.M., and Echols, E.E., "Design of Gravel Packs in Deviated Wellbores," JPT, January 1979, pp. 109-15.
3. Torrest, R.S., "Deposit Buildup During Gravel Packing with Viscous Polymer Solutions and Water," JPT, February 1983, pp. 325-28.
4. Shryock, S.G., Dunlap, R.G., and Millhone, R.S., "Preliminary Results from Full-Scale Gravel-Packing Studies," JPT, June 1979, pp. 669-75.
5. Shryock, S.G., and Millhone, R.S., "Gravel Packing Studies in a Full-Scale, Vertical Model, Well Progress Report," JPT, July 1980, pp. 1137-43.
6. Shryock, S.G., "Gravel Packing Studies in a Full-Scale Deviated Wellbore," JPT, March 1983, pp. 603-09.
7. Elson, T.D., Darlington, R.H., and Mantooth, M.A., "High-Angle Gravel-Pack Completion Studies," JPT, January 1984, pp. 69-78.
8. Schroeder, D.E. Jr., "Gravel Pack Studies in a Full-Scale, High Pressure Well bore Model," SPE Paper No. 16890, 1987.
9. Hodge, R.M., "Gravel Pack Studies in Deviated Wellbores," SPE Paper No. 10654, 1982.
10. Gurley, D.G., and Hudson, T.E., "Factors Affecting Gravel Placement in Long Deviated Intervals," SPE Paper No. 19400, 1990.
11. Forrest, J.K., "Horizontal Gravel Packing Studies in a Full Scale Model Wellbore," SPE Paper No. 20681, 1990.
12. Penberthy, W.L. Jr., and Cope, B.J., "Design and Productivity of Gravel Packed Completions," JPT, October 1980, pp. 1679-86.
13. Penberthy, W.L. Jr., "Gravel Placement Through Perforations and Perforation Packing for Gravel Packing," JPT, February 1988, pp. 229-36.
14. Wilson, D.J., and Barrilleaux, M.F., "Completion and Operational Considerations for Multizone Gravel Pack in Deep, High-Angle Wells," OTC Paper No. 6751, 1991.
15. Van Ballegooyen, J., Giap, T.K., and Seng, T.K., "Experience in Gravel Packing Long, Perforated Intervals in Deviated Holes," JPT, November 1983, pp. 2079-86.
16. Penberthy, W.L. Jr., and Echols, E.E., "Gravel Placement in Wells," JPT, June 1993.
17. Johnson, M.H., Ashton, J.P., and Nguyen, H., "The Effects of Erosion Velocity on Filter Cake Stability During Gravel Placement of Open hole Horizontal Gravel-Pack Completions," SPE Paper No. 23773, 1992.
18. Penberthy, W.L. Jr., Bickham, K. L., Nguyen, H., and Paulley, T.A., "Gravel Placement in Horizontal Wells," SPE Paper No. 31147, 1996.

The Authors