November 2003

Petroleum Technology Digest: Gravel packing through the shoe saves horizontal openhole job

Horizontal gravel packs have been achieved by placing additional gravel through the shoes, salvaging what otherwise would have been unsuccessful completions
By Petroleum Technology Transfer Council

Gravel packing through the shoe saves horizontal openhole job

At Southern California’s Wilmington field, the operator has successfully completed horizontal gravel packs by placing additional gravel through the shoe. In one case, gravel packing through the shoe salvaged what would have otherwise been an unsuccessful completion

Steve Siegwein, Tidelands Oil Production Co., Long Beach, California

A conventional, top-down gravel pack was planned in a recent, horizontally drilled well at California’s Wilmington field. When the job prematurely packed off at the top of the liner, the operator, Tidelands Oil Production Co., was left with a difficult decision. The firm could attempt to fish the liner and start over, try producing the well as is, or hope to place additional gravel through the shoe.

Most operators will not attempt to pack through the shoe, for fear of becoming stuck. The job was completed by gravel packing through the shoe, salvaging what otherwise would have been an unsuccessful completion. The well has produced 20 months, with current output averaging 180 bfpd, without sand problems.


A 9-7/8-in. hole was drilled while logging to a casing point at 3,230 ft, MD/2,676 ft, TVD. Final inclination at casing point was 5° less than planned. The top of the target sand was intersected at 3,164 ft, MD/2,663 ft, TVD, with 77.1° of inclination.

At 3,230 ft, 7-5/8-in. casing was set. After changing over to a sized, calcium carbonate, polymer mud system, the casing shoe was drilled out in a high-rate, build-oriented mode. This was an attempt to stay within the thin sand. An additional 7.9° of inclination was built within the first 89 ft below the casing shoe. Over a 62-ft length, average dogleg severity was 11°/100 ft, whereas Tidelands’ design parameter for horizontal gravel packing is less than 3°/100 ft.

The remainder of the interval was drilled to a final depth at 4,836 ft, MD/2,759 ft, TVD, with minimal doglegs. No commercially productive oil sands were found below 4,130 ft, MD. The well was plugged back using a bull-plugged inflatable packer set in open hole at 4,214 ft.


A high-viscosity polymer pill was spotted in the open hole. The liner consisted of 975 ft of 4-1/2-in. wire-wrapped casing, 90 ft of blank and a gravel packer with a sliding sleeve extension. The liner was fitted with two centralizers per joint and run with a 3-1/2-in., flush joint tubing stinger that was stung into the gravel pack circulating (GPC) shoe. The liner was run on 4-in. drill pipe, and the packer was hydraulically set.

Next, circulating through the GPC shoe, the high viscosity polymer pill was circulated out of the open hole. The gravel packing tools were then placed in the normal position, and gravel packing began at a concentration of 0.7 ppg added (ppa) using 9.0-ppg, filtered, NaCl brine water. The pack prematurely sanded out with 23 cu ft of gravel in place. 


The job was resumed by gravel packing through the shoe of the liner. The 3-1/2-in. stinger was run in and stung into the GPC shoe at 4,208 ft. Circulation was established down the work string, and out the end of the GPC shoe into the annular area between the wire-wrapped screen and open hole, Fig. 1.

Fig 1

Fig. 1. The L-233 horizontal open gravel pack (through the shoe) can be seen in this well schematic.

Initial circulating rates indicated restricted flow (as expected). The initial pumping pressure was 1,350 psi at 3.7 bbl/min. Altogether, 630 bbl of brine water were pumped with 0.4 to 0.5 ppa of gravel. The fluid return rate started at about 65% to 75% of the pumping rate, then slowly declined. It improved at times, then ultimately declined to less than 0.1 bpm. An additional 120 cu ft of gravel were placed, with a final pumping pressure of 1,250 psi at 3.3 bpm. 

The stinger was pulled from the GPC shoe, and the hole was reversed clean; no gravel was seen in the returns. For both jobs, 143 cu ft of gravel were placed around the liner (118% of calculated annular volume).


The premature sand-out probably resulted from a severe dogleg immediately below the 7-5/8-in. casing shoe. The amount of gravel placed during the initial gravel pack attempt was equal to 80 ft of 4-1/2-in. wire-wrapped screen in a 6-3/4-in. hole, plus 90 ft of 4-1/2-in. blank in 7-5/8-in. casing. There was a severe dogleg in the hole from 27 ft to 89 ft outside the 7-5/8-in. casing shoe. Some gravel may have passed beyond this point, but bridging began to take place very early in the job. 

Subsequent gravel packing through the shoe allowed a significant amount of additional gravel to be placed from the bottom up, resulting in a successful completion. Initial circulating pressures indicated that the wellbore was partially collapsed. Weak fluid returns in later stages of the job showed that some of the gravel was fractured into the formation and did not contribute to useful coverage of the screen.


Since the L-233 job, seven additional, horizontal gravel packs have been completed. Of these wells, four have had less than 100% gravel placement in the initial, top-down attempt. In all four cases, an attempt was made to place more gravel through the shoe. 

Three of the through-the-shoe gravel packs were in wells that had been steam-stimulated. The original completions were cased, perforated and steam-stimulated to consolidate sand around the perforation tunnels. All jobs consisted of 4-1/2-in. wire-wrapped screens (600 to 1,200 ft) inside 7-5/8-in. casing that had limited perforations (19 to 48) on the low side of the hole. In all three wells, top-down sand-out occurred with 40 to 45 sacks in place (25% to 50% of calculated amount), due to the high rate of fluid loss in the formation. 

The through-the-shoe jobs had varying degrees of success. In one well, an additional 45% of calculated volume was placed for a total of 90%. However, this is the only well that has exhibited a subsequent sand problem. In the well with the longest liner and the lowest initial percentage of gravel in place, circulation could not be established through the shoe (inclination at shoe is 76°). The third well packed off, with an additional 17% in place for a total of 70%. All three wells have been on production for eight to nine months, producing at rates of 720 to 1,530 bfpd.

The fourth well was a 1,590-ft, horizontal, openhole pack. The top-down job sanded out with 80% of calculated volume in place. An additional 15 sacks, or 7.5%, were placed through the shoe. This well has been onstream for four months and is producing 600 bfpd, sand-free. 

Experience gained since the L-233 pack indicates that most attempts to add additional gravel through the shoe are successful. However, none, (except for well L-233), resulted in a pack with over 100% of the calculated volume in place. Well L-233 may have been a special case, with all the gravel placed top-down located within the first 200 ft. However, the three cased hole jobs may not be indicative of the overall success rate that is achievable, due to severe lost circulation encountered while packing these high fluid loss wells.  WO



Steve Siegwein is chief development engineer with Tidelands Oil Production Co. in Long Beach, California. His primary responsibilities are drilling, re-drilling and remedial well work engineering. Previous experience includes 10 years with Union Pacific Resources in California and 912 years of Gulf Coast experience with Chevron, Houston Oil & Minerals and Union Texas Petroleum. He holds BS and MS degrees in petroleum engineering from the University of Louisiana at Lafayette. He can be contacted at:


Related Articles
Connect with World Oil
Connect with World Oil, the upstream industry's most trusted source of forecast data, industry trends, and insights into operational and technological advances.