Well Control and Intervention

Unique intervention safeguards platform after kick-induced gas broach

After a seabed gas broach occurred under an offshore platform, a relief well intersection technique protected it while replacing a severed production well.

John Wright, John Wright Company, Boerne, Texas

An unplanned drilling event escalated to a seabed gas broach beneath an offshore platform. Gas flowed uncontrolled for 10 hr, stopping after the open-hole annulus collapsed around the drill pipe. The operator formed an Incident Management Team (IMT) that successfully P&A’d the well.

Subsequent damage assessment revealed that another of the platform’s production wells was severed below the mudline, but above the tubing surface- controlled subsurface safety valve (SCSSSV). The IMT chose a novel relief well type of intersection technique, to both safeguard the platform and replace the severed well. Without the relief well, high flow potential gas below the SCSSSV posed unacceptable danger to the platform, with no practical way to intervene if it failed. Surface intervention options posed an unacceptable risk. With the relief well, the problem was safely and successfully isolated. The relief well was then turned into a replacement producer.

OVERVIEW

The platform’s operator had planned to deepen a depleted well to increase gas production. This well was on a 12-string, steel jacket platform in 11 m (36 ft) of water. By all measures, the field was mature, and the project was routine. Basic construction included pulling the completion; abandoning the well’s lower portion; cutting and pulling the 9-5/ 8-in. production casing below the 13-3/8-in. intermediate casing; sidetracking in 12-1/4-in. hole to the cap rock above the deeper, more productive target; set new 9-5/8-in. production casing; drill 8-1/2-in. hole through the new reservoir; run a 7-in. liner; and complete the well and put it onstream.

The project was executed with a self-erecting platform rig and tender assist barge. The operator’s well construction team pulled the completion, as planned, and abandoned the lower portion. When pulling the 9-5/8-in. casing, however, they were only able to pull 500 (1,640 ft) m of the planned 1,400 m (4,593 ft) to allow a kick-off below the 13-3/8-in. casing shoe.

This unplanned event created consequences that required risk assessment and option evaluation. Options included:

• Milling the 900 m of 9-5/8-in. (2,953 ft) casing 
• Recovering the slot and starting a new well 
• Milling a window at 500 m (1,640 ft) and drilling to target with additional casing string at 1,400 m (4,593 ft) and reduced tubing size 
• Milling a window at 500 m (1,640 ft) with a casing whipstock and drilling to the planned 9-5/8-in. casing depth with minimum kick tolerance.

The last option was the least expensive of the alternatives. However, it held the highest risk, if a kick was encountered during drilling of the 12-1/4-in. hole section. The team assessed that the sands that would be encountered in this section should all be depleted. Therefore, the risk of a kick was assumed negligible. Based on this assessment, this option was chosen, and project execution continued.

GAS BROACH

Drilling continued, as planned, until Christmas Eve, just prior to setting the intermediate 9-5/8-in. casing, when a gas kick was taken from a thin sand stringer thought to be depleted.

The shut-in casing pressure (SICP) was greater than maximum allowable annulus surface pressure (MAASP). An attempt was made to circulate out this kick using the driller’s method. However, partial losses occurred immediately and continued for the remainder of the circulation time. After 9 hr, gas bubbles were observed in the sea, increasing in intensity until the well broached to the seabed below the platform, Fig. 1.

Fig. 1. A gas broach occurred beneath the platform in 11 m (36 ft) of water.

The rig was abandoned, and the tender barge was winched off location without incident, injuries or pollution. The well flowed gas to the seabed for about 10 hr and then bridged, with no apparent major damage to the rig or platform, except that one wellhead and conductor had subsided about 14 in. An ROV survey around the platform indicated that a crater had been gouged out by the flow gas, with dimensions of 25m x 15m x 8m (82 ft x 49 ft x 26 ft) deep, Fig. 2.

Fig. 2. ROV investigation of the crater under the blowout platform.

After a civil engineering team proclaimed the jacket was sound, diagnostic wireline logs were run in the well’s drill pipe. They indicated that gas was cross-flowing in the lower portion of the annulus, between the normally pressured sand stringer and shallower depleted sands. The logs also indicated multiple annular bridges between the cross-flowing zones and the sidetracked casing window.

The IMT made the decision not to dynamically kill the cross-flow by pumping down the drill pipe. This logic was based on the low possibility that the bridge might dislodge during pumping, and the gas broach would restart during a failed kill attempt. The IMT subsequently plugged and abandoned the well’s upper portion by perforating the drill pipe and squeezing cement in sequential zones above the cross-flow, back inside the sidetrack casing window.

To mitigate cross-flow consequences, a replacement and pressure relief well was drilled from a recovered slot on the same platform. The target was about 50 m (164 ft) lateral proximity from, and parallel to, the cross-flowing reservoir. This well was completed and produced at high rates to deplete the reservoir pressure that would stop the cross-flow. This new well included an electromagnetic ranging bypass survey to fix relative downhole positions between the two wells. This allowed the parallel producer to be turned quickly into an intersection and hydraulic kill relief well, if needed.

KNOCK-ON DAMAGE TO AN ADJACENT WELL

While bringing the platform wells back onstream, unusual gas bubbling surfaced at sea level while attempting to pressure-up the well’s tubing. It subsided by 14 in. during the initial damage assessment. Diagnostics performed on the well, which included pumping tests on all annuli, lead impression block and downhole video, concluded that the tubing and protective casing strings had lost pressure integrity. They were open to the seabed just below the mudline.

The tubulars were assumed to have flow cut at the 26-in. casing shoe by high-velocity sand erosion, as gas expanded to the seabed around the conductors during the 10-hr broach (flow estimated at 5.3 MMcfgd). The IMT was unable to determine if the strings were completely severed or just cut, Fig. 3.

Fig. 3. Severed casing strings.

The 1,500-psi tubing pressure in this damaged well was being contained by the SCSSSV, with possibly dropped and buckled tubing, both in unknown condition. This situation left the subsided well with a questionable, single well control barrier on both the tubing and annulus, and no practical way to intervene, short of a relief well, if that barrier failed. Uncontrolled flow at the seabed under the platform was modelled, using the OLGA Well Kill software package, at blowout rates as high as 44 MMcfgd.

This would jeopardize the entire three-platform asset, including a second 12-well platform and production facility connected by bridges. This complex was a prolific gas producer. Leaving the well in a “time bomb” condition was unacceptable. However, conventional intervention posed a catastrophic risk of starting a broached blowout under the platform that would not bridge.

INTERVENTION STRATEGY

The IMT assessed several options to isolate the problem well (PW) with the following primary design criteria:

• Minimize risk to intervention personnel and the platform asset 
• Continue production from adjacent wells during intervention 
• Provide evidence that the well was isolated as the project completed.

Surface control options utilizing snubbing units or rotating heads and diverters were immediately rejected. They had a high risk of escalation, if the tubing or SCSSSV failed during intervention, and simultaneous production from the platform would not be allowed.

Three relief type, intervention well (IW) strategies were evaluated. The first option was to drill an IW and intersect the PW above the production packer, mill a notch in the casing and cement the tubing/ production casing annulus. With this accomplished, a slot would continue to be milled in the PW casing until the tubing was cut, then bullhead cement would be inserted into the tubing, isolating the reservoir. The IW would then be steered to the reservoir as a replacement well. A similar operation had been implemented in the Gulf of Mexico two years earlier, as part of a strategy to P&A and replace a single-caisson oil well knocked over by a shrimp boat.

The second proposed option was to drill an IW and parallel-intersect the PW below the 9-5/8-in. casing and attempt to mill a slot along the 7-in. liner. Then, tubing would be cut and dropped below the packer. A cement plug could be placed in the liner and squeezed into the tubing and perforations isolating the reservoir. An operation similar to this was successfully completed in Hungary in 2000, where a 21-ft slot was milled along a 7-in. liner from a relief well. The liner was re-entered from the relief well, and a cast iron bridge plug was set above the perforations as part of P&A operations.

The third proposed option was to drill an IW from the adjacent bridge-connected platform and intersect the PW along its production liner string, both above and below its perforated interval in the producing reservoir. After intersection was confirmed using electromagnetic homing-in technology, a liner would be set in the IW. The lower intersection zone would then be perforated through both strings, using oriented perforating guns.

With adequate hydraulic communication confirmed, the upper intersection zone would be perforated using the same technique. A re-settable test packer would be used to confirm circulation between the lower and upper perforations through the PW’s production casing. With this confirmed, a retainer would be set in the IW just above the lower intersection, and cement would be circulated down the drillstring, through the retainer, up the PW’s production casing and back into the IW’s annulus via the upper perforations.

The IW would then un-sting from the retainer and circulate out excess cement from its annulus to confirm cement placement through the PW. This would cement the PW tubing, perforations and production casing annulus between the lower and upper intersection depths. As a final assurance, a balanced cement plug would be set across the upper perforations and squeezed into the PW, pushing it up its tubing to a height above the production packer. With the intervention completed, the IW would be turned into a producer, replacing the isolated PW, Fig. 4.

Fig. 4. Perforating option, mechanical cross-section.

The IMT assessed each proposal against the design criteria and chose the third option as the most appropriate. In addition, the IW would be planned so that options one and two could be attempted as a contingency, if the perforating option proved unsuccessful.

DRILLING THE INTERVENTION WELL

The IMT managed the project. This team consisted of engineers and managers from the operator, as well as long-term, contracted service companies, including: Smedvig, Schlumberger, Baker, Red Baron and others. The specialist design and supervision team was managed by the John Wright Company and included: Well Flow Dynamics, Vector Magnetics and Scientific Drilling International.

The IW was spudded from an existing low-production well on the adjacent bridge-connected platform. This well was plugged, and tubulars were cut and pulled to allow sidetracking from its 13-3/8-in. casing, using a whipstock.

Fig. 5 shows the IW trajectory, which would closely track the target PW trajectory. The IW was directionally drilled toward the target, using conventional steerable motors and MWD tools. When the surveyed proximity between the two wells was 30 m (98 ft), electromagnetic homing-in tools were used to guide the relief well toward the target well casing.

Fig. 5. The intervention well (IW) trajectory closely tracks the target, problem well (PW) trajectory. The IW was directionally drilled toward the target, using conventional steerable motors and MWD tools.

The 9-5/8-in. casing on the IW was set parallel to, and about 3 m (10 ft) above, the target PW’s 9-5/8-in. casing. The IW intersected and then skidded along the target casing from 2,936 m (9,633 ft), MD, to 2,986 m (9,797 ft), MD – analogous to landing a plane on an aircraft carrier.

Because of the complexity of the attempt, maintaining the desired proximity between the two wells required multiple homing-in runs. Continuous, north-seeking gyro surveys were made in tandem with the homing-in runs. The upper intersection, perforating target, was chosen just below the PW’s 9-5/8-in. casing shoe.

The IW was purposely steered away from the target casing below this depth to maintain roughly a 4-m (13 ft) proximity when crossing the PW’s perforations, to avoid premature communication. The IW was then steered back to the lower intersection, where the wells touch again between 3,356 m (11,010 ft) and 3,388 m (11,115 ft) in impermeable shale below the reservoir.

The IW’s 7-in. liner was set and cemented at this depth. Then, the lower intersection zone was perforated through both strings using TCP guns with 0-5-0-355° phasing, oriented with a north-seeking gyro toward the target casing.

Hydraulic communication tests in perforations were excellent, with negligible pressure drop. Once hydraulic communication was confirmed, the upper intersection zone was perforated using the same technique, Figs. 6, 7 and 8.

Fig. 6. Upper intersection details, looking down the borehole.

Fig. 7. Lower intersection details, looking down the borehole.

Fig. 8. Details of tubing-conveyed perforating at upper intersection: 2,980 to 3,000 m, MD.

To prove that a sufficient circulation rate for cementing could be achieved between the lower and the upper intersection, a resettable test packer was set 5 m (16 ft) below the upper perforations. The mud was then circulated down the drill pipe through the lower perforations and up the lower part of the PW wellbore, out through the upper perforations and further up in the IW annulus. Circulation rates required for cementing were established with only limited additional, frictional pressure drop pumping through the two sets of perforations. After both well volumes had been circulated BU, a gas peak was recorded at surface, also indicating that communication to the PW wellbore had been achieved.

PRIMARY CEMENTING OF THE PW

After a proper communication path through the lower part of PW had been proven, a cement retainer was set using wireline above the lower perforations. To prove that the cement retainer had been properly set, the test packer was run down below the upper perforations, and pressure was applied to test the retainer setting. After tripping out with the test packer, a cement stinger assembly was run on drill pipe for connection to the cement retainer. Connection was observed at the expected depth with the required weight on pipe.

Prior to stinging in to the retainer, pump pressure at constant circulation rates was recorded. When pumping at the same circulation rates after being stung into the retainer, only limited additional pressure drop was observed, indicating that both sets of perforations were sufficiently open to achieve communication. The stinger was then pulled up 2 in. to close the communication path. The expected sudden increase in pressure was observed, proving a flowpath through the cement retainer.

To re-establish communication, the drill pipe with stinger was set down on the retainer again, and circulation rates were established. Ahead of the cement slurry, ester, chemical wash and mud push were pumped to clean up the PW. The cement slurry, with 100% excess volume and 10 hr of thickening time, was circulated down the IW drillstring, through the retainer, up the PW’s production casing and back into the IW’s annulus. To clean up the IW wellbore, several full circulations were performed, removing excess and contaminated cement slurry and spacers. This left cement up to the lower parts of the PW’s upper perforations. Spacers and excess cement came back as expected. No losses were observed during the cement job.

To cover the upper perforations that were left open by gravity during the cement program’s first part, and to further squeeze cement into the tubing, a balanced cement plug was spotted across the upper perforations. To avoid cement plug slump, a heavy, highly viscous mud pill was spotted from the cement retainer to 100 m (328 ft) below the upper perforations as a base for the cement. Prior to placing the cement, an injection test was performed, showing that communication still was available through the upper perforations and further up the tubing. A gas peak recorded at surface after circulating bottoms-up also indicated that this was true.

When the cement plug was in place, the drill pipe and stinger were pulled back into the 9-5/8-in. casing above the 7-in. liner top, circulated bottoms-up, and pressure was applied to squeeze 1.5 m³ of additional cement into the PW wellbore. Pressure was applied for 12 hr before bleed-back and drill-out of the cement in the IW could begin. After drilling passed the upper perforations, another pressure test was performed. The wellbore was pressured up to 7,000 kPa without any losses observed in the PW. The well was cleaned up to TD, completed and prepared for production. See a schematic of the basic cementing steps, Fig. 9.

Fig. 9. Lower perforations indicated that hydraulic communication (1) had been established between the wells for tubing cleanout in preparation for cementing. Cement was circulated into production casing (2) and squeezed into tubing (3) before the well was completed (4).

SUMMARY/ CONCLUSIONS

This technically challenging intervention was an industry “first.” It proves that today’s sophisticated ranging and directional technology can make exotic well connections possible and practical. In this case, tracking and steering through a very narrow window, with constant magnetic interference, to cement a damaged well eliminated the danger to a significant gas production platform.

Investigation of the intersection between the wells shows that a good cement plug can be achieved by using oriented TCP guns to perforate the well in two places, then circulate and squeeze cement through perforations.

Asset managers need to know and assess their well control-related and knock-on consequences in a systematic fashion. This is particularly critical in management of change scenarios, such as described in this article. Risk must be assessed independently for personnel and third-party safety, environmental damage, assets at risk, cost of control and recovery, business disruptions and reputation damage.

Furthermore, senior drilling supervisors and engineers need to be trained for underground cross-flow recognition and response actions. It might have been possible to have prevented the broach, if heavy kill fluid had been pumped immediately, increasing the hydrostatic pressure on the reservoir between the bit and the loss zone at the casing window, using a sandwich kill technique.

One-trip, continuous gyro surveying and electromagnetic ranging in the same pressure barrel was successfully implemented for the first time, saving several days of rig time by avoiding round trips to make gyro surveys at the bottom of the well. Furthermore, the upper intersection was achieved without problems, providing more than 40 m of less than 0.1-m proximity (edge to edge) between the two wells. No problems were experienced when drilling along-side the 9-5/8-in. and 7-in. liner casing strings, even with several collar connections and a high density of centralizers along the PW shoe track.

While drilling at 3,388 m, alongside the PW for the lower intersection, losses occurred via a suspected hole knocked into the 7-in. liner by the bit while drilling in the oriented mode toward the liner. This caused a short increase in project duration, but it was resolved by running an external casing packer and cementing stage tool to isolate the hole from the IW’s 7-in. liner string cementation.

In addition, the upper and lower intersections were perforated with an oriented TCP gun with excellent results. After testing communication between the two intersections, plugging proceeded, as per textbook. The final squeeze plug held for 12 hr. A final pressure test after drilling out the IW cement plug indicated a success.

The IW was successfully turned into a replacement well. A self-erecting tender assist rig proved to be a good choice for the intervention. It provided rapid escape capabilities and more than adequate storage capacities and operational capabilities. The total project took 76 days. 

THE AUTHOR

John Wright, formed the John Wright Company in 1989, to provide blowout control engineering design and specialty services related to relief wells and underground blowouts. Since then, he has designed and supervised 33 relief well and borehole intersection projects worldwide. Mr. Wright cofounded Well Flow Dynamics in December 1991, and also managed the field operations of Vector Magnetics, a casing detection firm, in 1991 and 1992. He has authored 15 relief well and blowout-related industry publications. Prior to 1989, he worked for Eastman Christensen and Schlumberger. He holds a BS degree in mechanical engineering from Texas A&M University. Email: jwright@jwco.com