Selective horizontal gravel packing on the Alba Field

Intra-reservoir shales complicate completion

Alba is an Eocene heavy oil field located in the Central North Sea, that currently produces about 100,000 b/d of 19 degree API oil. The field consists of a relatively homo genous, unconsolidated sandstone reservoir with high permeability (about 3.5 darcy) and underlying water. Therefore, completion design is heavily influenced by both water coning and sand control requirements.

The first 13 horizontal producers were completed with pre-pack screens; however, in response to a number of sand control problems, a change was made in 1996 to the newer type of all-metal screen. Eight producers currently have been completed with these Excluder screens.

Due to the history and severity of Alba's sand control requirements, additional forms of sand control have been investigated. Horizontal gravel packs have proved successful in other parts of the world, but have generally been limited to low rate wells with clean sand intervals. While Alba's sand is very clean, intra-reservoir shales are routinely encountered by the horizontal wells, hence wells typically have a net-to-gross of 70% over the completed interval.

Shale stability

Alba's experience has shown that short-reach wells can be drilled and completed with water-based mud, despite drilling long intervals of fairly reactive shale. However, to perform a horizontal gravel-pack, it would be necessary to displace the open-hole section to brine.

To address this concern, a series of shale stability tests were performed to identify chemical additives for the brine, to inhibit shale swelling and provide a chance of gravel packing a well with some exposed shale. Gravel-packing a well with unstable shale would run the risk of low productivity, due to shale entrainment in the pack, and have a significant risk of premature bridging, leaving some of the interval unpacked.

In 1997, the decision was made to attempt a horizontal gravel-pack. However, it took until February 1998 to find a suitable candidate, well A25, a re-drill of A8z, necessitated by a pre-pack screen failure. A 3D computer modeling technique was used to select the gravel size, based on a probabilistic pore-throat size distributions of different gravel sizes. Size 40/60 gravel was selected, instead of 30/40, suggested by Saucier's rule, as a compromise between the risk of plugging the gravel-pack and preventing fines production.

The open-hole section of A25 was drilled with 10.8 ppg Baradril-N (a water-based mud weighted with calcium carbonate) using a rotary-steerable drilling assembly. Due to geological constraints, the open-hole section turned out to be relatively short for Alba, at 825 ft long, with 677 ft of sand and 148 ft of intra-reservoir shale at the toe of the well. Also, to maximize the height of the well above the oil-water contact, the well path built along its length reaching 94 degrees inclination at the toe.

Completion

After drilling to TD, a clean-up assembly was run to displace the well to brine, leaving a thin layer of calcium carbonate filter-cake held in place by hydrostatic pressure, to prevent losses as the well was approximately 800 psi overbalance. However, some shale stability problems were encountered and the cleanup assembly hung up 61 ft into the shale section, preventing the cleanup assembly from reaching TD.

The sand control screens and gravel pack assembly were run in brine, significantly reducing the risk of plugging the screens while running in hole. Unfortunately, shale swelling prevented the screens reaching TD, so the screens were set 130 ft off bottom.

Losses were first encountered when drilling the toe of the well and these continued at up to 15 bbl/hr throughout the clean-up and screen running phases.

Since these losses coincided with drilling the intra-reservoir shale, this was attributed to instability along the sand/shale interfaces. Unfortunately, the loss rate increased, whilst setting the gravel-pack packer and pickling the drill pipe. When it came time to start the gravel-pack operation, no returns were obtained despite pump rates up to 6 bbl/min. This would have precluded gravel-packing, as such losses would have caused the gravel to bridge-off prematurely. To cure these losses, a 10 bbl pill of drilling mud was spotted into the open-hole section. This was successful and allowed the gravel-pack to proceed.

The pack was pumped at 6.5 bbl/min with 1 ppg gravel and a return flow rate of 4-5 bbl/min. In total, 130% of the theoretical open-hole volume of gravel was pumped, of which about 10% was reversed at the end of the job. The final screen out pressure was 2,000 psi at 3 bbl/min, indicating that the gravel was packed 20 ft above the top of the screen. Therefore, once the loss rate had been stemmed, the operation went reasonably well and the entire was successfully packed.

The initial productivity of the A25 well is ?100 bbl/day/psi, which is acceptable, considering the relatively short sand interval. The specific productivity index of 0.16 bbl/day/psi/ft compares favorably with other Alba water-based mud wells. This productivity index also lies within Alba's ideal range of 100-200 bbl/day/psi, which ensures both good production rates and inflow over the complete open-hole section.

A second horizontal gravel-pack was planned for well A26; however, this well intercepted over 1,400 ft of shale, which was not sufficiently inhibited to risk attempting the gravel-pack. The decision to gravel-pack must therefore be made on a well-by-well basis. However, where there are stable shales, there is a good chance of achieving a good gravel-pack and improving the overall completion.