J. Bret Lane, J. Parks Wesson
Souther California Gas Co.
Los Angeles
Magnetic ranging surveys and directional drilling technology accurately guided a replacement well bore to intersect a leaking gas storage well with casing damage.
The second well bore was then used to pump cement into the original leaking casing shoe.
The repair well bore kicked off from the surface hole, bypassed casing damage in the middle of the well, and intersected the damaged well near the casing shoe. The repair well was subsequently completed in the gas storage zone near the original well bore, salvaging the valuable bottom hole location in the reservoir.
Southern California Gas Co. (SoCalGas) is the local distribution company for Southern California, serving more than four million customers. To augment its pipeline supply system, SoCalGas operates five gas storage fields in the Los Angeles area.
The five fields are depleted oil or gas reservoirs that have been converted to gas storage. SoCalGas began operating its first storage field in 1938 and acquired its latest field in 1975. This field, Honor Rancho, is about 60 miles north of Los Angeles and has a 17.5 bcf working gas volume serviced by 36 wells.
A storage well at Honor Rancho, WEZU 30, developed a shoe leak. During a repair attempt, problems developed in the old casing, preventing proper leak repair.
SoCalGas determined that the safest method to handle this deteriorated well would be to plug it and replace it with a new well at the same location. The replacement well would use a magnetic ranging tool for precise directional control.
The new well bore was planned to sidetrack from the bore of WEZU 30, drill alongside, and then intersect the original well bore in the shale overlying the storage zone. The gas channel from the storage zone would then be sealed properly.
This method would prevent the loss of storage gas, and it would prevent a potential underground blowout that could permanently damage the integrity of the storage field.
WELL BACKGROUND
Well WEZU 30 was directionally drilled and completed as an oil producer in 1961. In 1968, the well was worked over to repair a casing leak at 7,500 ft, and some casing damage was found. The casing section was repaired by milling through the damage and setting a 4-in. casing patch across the interval.
In 1976 during a workover to convert the well for gas storage, SoCalGas found that the casing damage had increased. Most of this damage occurred near the doglegs created where whipstocks were used to drill the well directionally. After the casing patch was pulled, repeated milling runs were made to repair the damage. A 4-in. casing patch was run from 7,200 ft to 8,410 ft and cemented in place.
A casing shoe leak was found during routine monitoring checks with temperature surveys and noise logs. In January 1991, a workover rig was brought in to repair the leak by squeezing cement into the channel.
If the leak could not be repaired, the well would be abandoned. During the workover, an isolation plug that had been placed in the packer could not be retrieved. The small-diameter casing patch prevented the plug from being easily retrieved. After repeated attempts to recover the plug, SoCalGas decided to mill out the cemented casing patch. This operation would then allow the plug and packer to be retrieved and the leak to be properly repaired.
The casing patch was milled from 7,220 ft to 7,500 ft. A dogleg 7 at 7,500 ft caused the mill to exit both the 4-in. inner casing and the 5-1/2-in. outer casing and enter the formation.
Repeated attempts to reenter the casing with small mud motors and steering assemblies failed. The small internal diameters of the 5-1/2-in. casing and 4-in. casing patch and the dogleg severity limited the selection of directional equipment needed to re-enter the casing.
After trying to re-enter the old casing for more than 2 months, SoCalGas decided to suspend operations and explore other options available with current technology. A gyro survey was run from 7,500 ft to the surface to locate the old well bore exactly. A cement plug was then placed at 7,000 ft, and the workover operation was temporarily suspended.
PLANNING
A literature search and discussions with several directional companies suggested that the technology was available to sidetrack the well and intersect the casing below the casing patch with tools developed for drilling relief wells for blowouts. SoCalGas contacted two companies that provide magnetic ranging tools, and one company indicated it was possible to intersect the well and re-enter the casing.
As the plan developed, it became evident that many specialized contractors would be needed. Close coordination and clear lines of responsibility would be required for a successful operation.
Separate meetings were held with representatives from the magnetic ranging tool company and members of the directional drilling company's relief well team, the measurement-while drilling (MWD) company, and the fishing tool company. These meetings provided essential information on the available services, the type and capability of the required equipment, the potential problems, and the probability of success.
These meetings led to a plan to salvage the upper section of the existing well and to plug the well to a specified depth. A new well bore would then be kicked off and directionally drilled beside the old well to a defined target depth. Magnetic ranging tools would guide the new hole into the old casing. Then a short section of the casing could be milled to abandon the well.
In case the first attempt to intersect the old casing failed, the intersection point was planned high enough above the storage zone to allow additional attempts.
The John Wright Co., relief well specialists, helped prepare and implement the intersection program. This company was familiar with operating magnetic ranging tools and had experience in directional drilling.
The specialist from John Wright helped develop a well intersection plan, interpret the data from the magnetic ranging tool, and adjust the directional plan as necessary to ensure the proper target intersection. The magnetic ranging tool, Wellspot, from Vector Magnetics Inc. was chosen based on industry recommendations and on the familiarity of the tool by the specialist from the John Wright Co.
MAGNETIC RANGING
The Wellspot magnetic ranging tool consists of a downhole current injector and a magnetic field sensor sonde separated by an insulated bridle assembly with an adjustable length. The tool is run in the open hole of the relief well on a conventional seven-conductor wire with an open hole logging unit.
Once the tool is at the bottom of the relief well, the current injector emits an alternating current. This current flows through the earth and concentrates on the target casing, generating an electromagnetic field with flux lines perpendicular to the target casing.
The sonde senses the orientation of these lines and their relation to the earth's magnetic field and the tool's high side. Subsequent surveys are made at different depths. The data are then used to determine the direction of the target casing in relation to the relief well. The distance between the target and the relief well is proportional to the strength of the signal.
Once the location of the target is established at one depth, an intersection plan is made, and drilling can proceed. As the relief well approaches the target, additional logging runs are made to confirm and refine the intersection plan for the new depths.
INTERSECTION
The preliminary plan was to sidetrack the relief well and drill it alongside the original well. This relief well would intersect and enter the target casing at 8,500 ft. This intersection point was below the bottom of the casing patch at 8,410 ft and above the packer at 10,100 ft. This range provided an additional 1,600 ft to intersect the target if the first attempt failed.
Once the relief well entered the casing, the packer would be milled and the shoe leak properly repaired by perforating the channel and squeezing in cement. The plan called for the relief well to intersect the target at a 3 angle of incidence. Between six and eight Wellspot runs were estimated for the target intersection.
A review of the original 1961 single-shot surveys showed that an incorrect magnetic declination was used to calculate the well's position. Once corrected, these surveys closely matched the gyro survey from the earlier workover operation. Because there were no gyro survey data below 7,500 ft, the single-shot data from 1961 had to be used from that point down. The close match between the two methods indicated the target was well defined.
The surveys provided a predictable target for the relief well. Because the well location was known, fewer Wellspot runs would be required than if no surveys were available.
REPLACEMENT WELL
A drilling rig moved onto well WEZU 30 on Aug. 4, 1991. Stuck pipe logs indicated the upper 7,000 ft of 5 1/2-in. casing was partially free and could be removed from the well.
However, only the upper 2,953 ft of casing could be pulled. The well was plugged back to 2,400 ft and sidetracked with a 9/s-in. bit. The well was directionally drilled following the directional plan to 7,247 ft using a downhole mud motor and an MWD tool.
A Wellspot run and a gyro survey were made at 7,247 ft. The gyro data from the relief well closely matched the MWD data, confirming the MWD accuracy. This correlation implied there was no magnetic interference between the MWD and the target casing.
The gyro data from the relief well, along with the original 1961 single-shot survey of the target well, determined the direction and distance between the two wells. These data closely matched the Wellspot results. The accurate data from the magnetic ranging tool indicated that the old casing could probably be intersected on the first attempt.
The plan was revised to intersect the target casing at 10,140 ft in a shale immediately below the packer at 10,100 ft (Fig. 1). This intersection point (not drawn to scale) was 60 ft above the storage zone at 10,200 ft.
This revision was based on the accuracy of the Wellspot tool from the previous survey and the belief that the well could intersect a much tighter target of 60 ft instead of the original 1,600 ft window. The revised plan eliminated the need to mill out the packer to reach the storage zone. Also, if the well could not be re-entered after it was intersected, the shoe leak could still be repaired properly by squeezing cement into the zone.
Drilling continued with the mud motor to 9,789 ft. Two Wellspot runs and another gyro survey were made in this interval. The directional plan was refined and drilling continued.
UNPLANNED INTERSECTION
On Aug. 26, 1991, the casing was contacted unexpectedly at 9,860 ft (Fig. 2). A review of the MWD data from the relief well and the survey from the original well indicated there should have been 4 ft between the relief well and the casing at that depth.
However, the mud motor was stalling, and metal picked up in the flowlines indicated the bit had hit the casing. A Wellspot run confirmed that the bit initially hit the casing at 9,848 ft and departed at 9,879 ft. The angle between the relief well and the severed casing was 3.
Several attempts to re-enter the target casing were unsuccessful because the 3angle of intercept between the relief well and target well was too severe to allow entry into the target casing.
Two problems contributed to the unplanned intersection:
- The magnetic interference from the target casing caused inaccurate readings from the MWD tool. Although MWD readings at the time did not indicate interference, subsequent gyro surveys confirmed the MWD was incorrect through this interval.
- Although the 1961 single-shot directional surveys were fairly accurate higher in the well, they were extremely inaccurate down in the target intersection area.
The previously well-defined target was now questionable because the old single-shot surveys down the well were thought to be of little value. Thus, the plan changed to rely completely on the magnetic ranging data. Several more Wellspot runs would be required for target intersection.
Drilling continued to 10,090 ft, where a sixth Wellspot survey was run. The survey indicated that a high dogleg would be required to intersect the casing, making it virtually impossible to reenter the casing.
PLUGBACK
The well was plugged back to 9,500 ft and kicked off for another target attempt.
The 9 7/8-in. well bore had been open for about 60 days. Drilling experience in this area has shown that well problems will typically begin 30-45 days after drilling starts.
After the first intersection well bore was plugged back, the work string began to encounter severe sticking problems in the well bore. Thus, 7-in. casing had to be set and cemented to 9,023 ft. The relief well was sidetracked, and drilling continued.
After 7-in. casing was run, a plan was modified to intersect the well at 10,100 ft at a l angle of incidence. The decision to lower the angle from 3 was based on the unsuccessful attempt to reenter the well at 9,860 ft.
The lower angle of incidence required a slower approach to the target, allowing easier path corrections if necessary. Also, if the relief well bypassed the target, the direction could be corrected to allow another intersection attempt.
Drilling continued to 9,910 ft where another Wellspot survey was run. However, the lack of-electrical continuity in the 5 1/2-in. casing, which had been completely severed from 9,858 ft to 9,868 ft, made it difficult to determine the distance to the target. The relief well was incorrectly estimated to be 1 1/2 ft from the target casing. The plan was revised, and contact was expected at 10,012 ft.
The motor began to torque up and stall at 9,912 ft, indicating contact with the target casing.
A specially designed flatbottom june mill was used to mill the casing from 9,912 ft to 9,934 ft. This depth provided an adequate point to squeeze cement and abandon the lower section of the well (Fig. 3).
The bottom-hole location of the old target well in the storage zone was critical for proper operation of the reservoir. The relief well was then plugged back to 9,023 ft and redrilled to 10,000 ft, 40 ft north of the previous bottom hole location. The well was then completed as a regular gas storage well (Fig. 4).
RECOMMENDATIONS
A repair operation of this nature is quite rare. Most wells can be abandoned if problems are encountered at shallow depths because the producing fields are on a constant pressure decline. In gas storage fields, however, a leak around the casing shoe must be repaired so the field can continue to operate.
Although this operation was extremely expensive, it was a promising application of available technology. This technology is a viable option for gas storage operators to repair leaking wells instead of abandoning them.
A number of factors should be reviewed before an operator considers this type of operation:
- The technology is available to drill and intersect old wells for proper repair. This technology need not only be used for large blowouts.
- The initial planning of all the parameters associated with drilling the relief well is critical. The engineers should know the geology at and above the target, the drilling characteristics of the formations to be encountered, and the accuracy of target well survey data.
- The selection of the relief well team is crucial. The team members should include the drilling operator, relief well specialist, directional company, bit company, and fishing tool company (for the proper milling tools). The project leader must establish clear lines of responsibility for all of the team members.
This type of operation normally has several experts on location. Too many people giving orders can lead to confusion or conflict. In addition, the workover is complex and will probably be a high-profile operation. This pressure can lead to a stressful work environment. The proper delineation of roles up front can reduce or eliminate some of these problems.
- If single-shot directional surveys from the target well are used, their accuracy should be verified with gyro surveys, if possible. If the surveys are not verified, the single-shot data should only be used as a guideline, and additional magnetic ranging tool runs should be planned.
For this well, the accuracy of the old surveys in the shallow portion of the well misled the team to believe these surveys were accurate deep in the well. The inaccurate old single-shot data led to the unplanned intersections of the target casing.
Subsequent magnetic ranging tool surveys showed the old single-shot surveys to be several feet off.
- The workover team should use as many magnetic ranging tool runs as needed. Although initially expensive, an additional two or three ranging tool runs are cheaper than plugging back and redrilling a well because of a missed interception point.
- Certain unknown conditions may influence magnetic ranging accuracy. In this well, the strong magnetic poles generated from the severed casing resulted in an incorrect estimate of the distance to the target.
The team should examine the previous well bore trajectory and angle of intercept. Because the existing well trajectory has a direct influence on the new well trajectory, it is critical that the relief well be planned to intersect the target casing at approximately a 1 angle of incidence if a long section of target casing is to be milled.
BIBLIOGRAPHY
- Flak, L.H., and Goins, W.C. Jr, "New Relief Well Technology is Improving Blowout Control," World Oil, December 1983, pp. 57-61.
- Grace, R.D., and Kuckes, A.F., "Operations at a Deep Relief Well: The TXO Marshall," SPE paper 18059, Society of Petroleum Engineers Annual Technical Conference and Exhibition, Houston, Oct. 2-5, 1988.
- Morris, F.J., Waters, R.L., and Roberts, G.F., "A New Method of Determining Range and Direction from a Relief Well to a Blowout Well," SPE paper 6781, Society of Petroleum Engineers Annual Technical Conference, Denver, Oct. 9-12, 1977.
- Voisin, J.A, Quiroz, G.A., Pounds, R., Wright, J., and Bierman, K, "Relief Well Planning and Drilling for SLB-5-4X Blowout, Lake Maracaibo, Venezuela," SPE paper 16677, Society of Petroleum Engineers Annual Technical Conference and Exhibition, Dallas, Sept. 27-30, 1987.
- Warren, T.M., "Directional and Proximity Log Analysis of a Downhole Well Intersection," SPE paper 10055, Society of Petroleum Engineers Annual Technical Conference, San Antonio, Oct 5-7, 1981.
Copyright 1992 Oil & Gas Journal. All Rights Reserved.
