Well Control

How a snubbing unit was used to regain control of a subsea well

Snubbing equipment on a jackup, a specially fabricated high-pressure riser and the original subsea BOPs were used to kill and plug a problem well

David A. Barnett, Engineering Manager, Wild Well Control, Inc.; and D.C. Tyagi and A. K. Mehra, Oil & Natural Gas Corp., Ltd.

Well B-24-2 was drilled in the Bombay High region offshore India. Even though water depth was 266 ft, a drillship was used for drilling since underlying soil instability in the area might allow a “punch-through” if a bottom-supported rig were used.

The well encountered a kick while displacing mud prior to completion operations. Complications, which arose while circulating the kick, made it necessary to shear the drill pipe. Characteristics of the pressure influx path and other factors created a situation in which further application of conventional well control techniques was not advisable. A thorough evaluation of risks versus the probability of success was carried out for several possible control techniques. Based upon these evaluations, it was decided that use of a snubbing unit, along with a specially fabricated high-pressure riser system, would be the most prudent course of action.

Subsequently, the well was successfully controlled through operations performed according to a plan developed prior to project initiation. Numerous technical and logistical challenges, which were dealt with during the project, are discussed below.

WELL PLAN

An expendable exploratory well, B-24-2 is located in the southern section of the Bombay High offshore area. Well initially was targeted to a depth of 9,892 ft, which was later revised to 10,390 ft. Based on available pressure data, the well was designed with the following casing program:

RFT results from a nearby well indicated a maximum pore pressure of 14.2 ppg equivalent mud weight (EMW). Thus, the mud program called for a maximum 14 to 15 ppg mud weight at TD. Anticipated formation tops and the well plan are given in Fig. 1.

Fig. 1. Well B-24-2 was the third exploratory well on the structure. Both earlier wells did not reach target depth due to pressure activity and complications. After setting and testing a 5-in. liner at 10,384 ft, an open-ended string was run to displace mud with seawater in preparation for testing casing. An influx was observed while displacing the mud, which eventually escalated to a well control situation.

The B-24-2 was the third exploratory well on the structure. Both earlier wells did not reach target depth due to pressure activity and complications encountered at 9,800 ft in the first well and 10,128 ft in the second well.

DRILLING OPERATIONS

An ONGC drillship was used to drill the B-24-2 well. Drilling of the 12-1/4-in. hole section was suspended at 7,697 ft due to severe weather conditions (cyclone). The well was temporarily abandoned with the drillstring hung off on the pipe rams and the shear blind rams (SBR) closed.

After conclusion of monsoon season, drilling resumed with another ONGC drillship. Drilling proceeded to a depth of 9,374 ft, where 9-5/8-in. casing was set and cemented.

Data indicated that permeable, high-pressure sands coexisted with coal beds, and these were expected to cause lost circulation problems below the 9-5/8-in. shoe. Drilling of the 8-1/2-in. hole required an optimum hydrostatic balance and close monitoring of drilling parameters to minimize the possibility of well control problems. The first coal bed formation was encountered at 10,302 ft, and a second coal bed was found at 10,325 ft. Both sections were drilled without incident.

A drilling break occurred at 10,341 ft, and the well was shut-in after a positive flow check was observed. SIDPP of 200 psi and SICP of 250 psi were recorded. The shut-in pressures suddenly dropped to 0 psi, indicating a loss of mud into the suspected thief zones. The well alternated between 53 bbl/hr losses and self flow, which resulted in SICPs as high as 800 psi. The well was eventually controlled with 16.1 ppg kill mud after repeated LCM and cement pills. In view of the uncertainties expected below 10,341 ft, a 7-in. liner was set with the shoe at 10,322 ft.

Drilling continued below the liner to the target depth of 10,390 ft with 16.1 ppg mud. Insufficient hole fill-up was observed while pulling out of the hole with the drilling assembly. The well was shut-in and a SICP of 450 psi was observed. This influx was attributed to the loss of equivalent circulating density (ECD) and/or swabbing caused by the 4-3/4-in. drill collars in the 6-in. hole. The well was eventually controlled via an off-bottom kill (29 stands) using 16.8 ppg kill mud. The drillstring was run to bottom and the mud weight was reduced to 16.5 ppg. A 5-in. liner was set and cemented with the shoe at 10,384 ft. The liner and liner top were both tested.

INITIAL WELL CONTROL INCIDENT

An open-ended string, consisting of 2-7/8-in. and 3-1/2-in. tubing and 5-in. drill pipe, was run in the hole to TD in order to displace the mud with seawater in preparation for testing the casing.

An influx was observed while displacing the mud. Drill pipe was hung-off on the upper pipe rams and immediate steps were taken to circulate the influx with 16.5 ppg mud. While circulating gas from the wellbore, the de-gasser vent line ruptured causing dangerous accumulations of gas to form around the rig floor.

Well control operations were suspended while an auxiliary overboard vent line was quickly installed. However, gas from the temporary overboard vent line ignited, and due to crew safety concerns, fail-safe valves had to be actuated, shutting in the well. This resulted in a SIDP of 2,500 psi and SICP of 4,500 psi.

To compound problems, a grease nipple blew off of the low-torque valve that was installed on the drillstring. Efforts to close the full-opening safety valve (FOSV) were unsuccessful. Therefore, drill pipe had to be sheared at the subsea BOP stack. Shut-in pressure was recorded at 5,400 psi.

Efforts to control the well by lubricating mud were carried out for about 10 days. The annulus pressure (monitored below the pipe rams) was eventually reduced to 1,200 psi. The pressure recorded under the SBR cavity was 3,500 psi when monitored via the upper kill line (UKL). Fig. 2 shows a schematic of the subsea BOP stack at the time of these operations.

Fig. 2. Schematic of subsea BOP stack at the time the well control problems developed.

An attempt was made to bullhead kill mud down the 5-in. by 9-5/8-in. annulus through the lower choke line (LCL) while monitoring drillstring pressure on the UKL. A maximum surface pressure of 4,300 psi was applied to the annulus, while no change was observed on the drillstring. A similar attempt was made to bullhead mud down the drillstring through the UKL while monitoring annulus pressure. A maximum of 6,050 psi was applied, but annulus pressure remained steady. An insignificant volume of mud was injected during this procedure.

The behavior of the well during the kill attempts led the well control team to the following conclusions:

• Bridging in the annulus was indicated by the lack of communication between drill pipe and annulus (bridge most likely to be below the 5-in. liner top).
• Source of pressure in the annulus was probably a leaking liner top.
• Source of pressure in the drillstring was probably from communication with a high-pressure, permeable zone below the end of the tubing string.
• It was not possible to bullhead kill weight mud with reasonable and safe surface pressures.

A number of safety and reliability-related issues were evaluated to determine the best forward plan. Among the issues discussed were:

• Continued pumping operations would involve pressures close to test pressure of the subsea BOP stack.
• Continued cycling of the fail-safe valves could lead to serious complications and/or catastrophic failure.
• The probability of success using lubrication and/or bullheading techniques was very low.
• The disruption of a bridge in the annulus could result in very high sustained surface pressures and/or underground flow.

After thorough consideration of the various possibilities, it was agreed that the safest, most reasonable plan with the highest probability of success was the application of snubbing equipment and related services. The snubbing unit would be used to tie-back the drillstring and provide a means to establish circulation at or near bottom.

The monsoon season in the Arabian Sea was approaching, and since it was not be feasible to initiate well control operations immediately, the project was scheduled for the post-monsoon season. The well was temporarily abandoned in its current condition. A check valve was placed in the hydraulic line connected to the closing chamber of the SBR. This was done to trap the closing pressure on the SBRs while preparations for the snubbing project were made. The hydraulic connector below the upper annular was unlatched, the drilling riser and lower marine riser package (LMRP) were removed and the drillship was released from location.

WELL CONTROL PLANS

Thorough planning is necessary for the safe, timely completion of any major project. However, several factors made precise planning an absolute necessity for the B-24-2 project. A number of individual tasks were identified and subsequently used to develop a scope of work and schedule for the project. Some of the major items included:

• Most suitable rig and safeguards against punch-through risk
• Rig positioning over subsea wellhead
• Best method/ equipment to tie-back the subsea BOP stack to the jackup rig
• Additional BOPs, pressure control equipment and configuration
• Detailed equipment and services list
• Mobilization plan.

Rig selection and punch-through. The potential for a punch-through occurrence made proper rig selection a critical component of the planning stage. An intense evaluation of all movable offshore drilling units (MODU) located in the immediate area was undertaken. The evaluation concluded that the most suitable rig for the project was the ONGC jackup Saga Shakti. ONGC and third-party personnel inspected the designated vessel.

A number of modifications were specified to prepare the rig for the well control operations. Deck load characteristics were addressed by removing all unnecessary equipment. A detailed analysis of loads that would be imposed during snubbing was performed. These loads included both static equipment weight and dynamic loads that would occur during operations (i.e., tensioning riser, pulling on pipe, etc.).

Rig positioning. Placement of the jackup rig over the subsea well was critical. The possibility of impacting the pressurized wellhead could have catastrophic consequences. Thus, extraordinary planning was done to minimize the possibility of such an occurrence. The positioning plan was developed in conjunction with, and approved by, the Warranty Surveyor. A summary of the positioning plan follows:

• Attach marker buoy to subsea BOP.
• At 100 m, pin down legs and set four anchors per pre-determined pattern, start moving in using anchors.
• At 50 m, install side scan sonar (starboard/aft spud can) and acoustic tracking system (transponder on wellhead, transducer on port/aft spud can).
• At 20 m, interchange transponder and transducer.
• Position rig to within 6 2 m and jack up to zero air gap.
• Check position relative to well with diver, adjust if/as necessary.

Once positioned, the rig was pre-loaded per water depth, environmental forces and calculated variable deck load and hook loads. The rig was then proof-loaded to capacity. Fig. 3 shows the position of the Sagar Shakti relative to the wellhead.

Fig. 3. Placement of the jackup rig over the subsea well was critical since the possibility of impacting the pressurized wellhead could have catastrophic consequences. Anchors, side scan sonar, transducers and transponders, and divers were used to position the Sagar Shakti relative to the wellhead.

Subsea tie-back. The 266-ft water depth made it necessary to use specialized equipment for reconnecting to the subsea BOPs. The riser between the subsea and surface BOPs would have to be capable of withstanding severe combined stresses caused by pressure, tension and bending moments. Since ordinary API flanges are normally de-rated when this type of combined stress is applied, a specially fabricated, purpose-built riser system would be used.

The major components of the high-pressure riser system are standard API 11-in., 10,000-psi spool sections with Steel Products Offshore (SPO) “Compact” flanges. The SPO flanges are specially designed to withstand the combined loading that would be anticipated on the B-24-2 project. In fact, these flanges are designed to maintain a pressure seal under stresses that would cause a failure in the 11-in., 10,000-psi spool body.

In addition to the spool sections, the riser system includes a specialized spider support system for installation on the rig floor, clamps for supporting auxiliary choke and kill line sections and a tension ring for lateral support from the rig structure.

BOPs and pressure control equipment. As with any snubbing intervention application, the BOP and pressure control equipment configuration is critical. In addition to being adequate for the anticipated pressures and fluids, the BOP system must be designed to establish a high-pressure (maximum working pressure) seal on all tubulars in the wellbore and allow operations to proceed with the surface pressure present. This includes the installation and removal of specific tools, considering their length, diameter and shape. While allowing certain operations to proceed, the snubbing BOP arrangement must also provide an extraordinary measure of reliability and redundancy.

The BOP system (Fig. 4) was designed with these factors in mind and to be independent from the existing subsea BOPs. BOPs were arranged to accommodate all pressure testing requirements that are associated with well control operations, as per industry standards.

Fig. 4. The snubbing BOPs must provide an extraordinary measure of reliability and redundancy. Separate choke and kill lines were installed to allow for extensive circulation. Dual valves were installed on the subsea drilling cross for safety. Choke and kill lines were fabricated of 5-in. drill pipe and connected to the riser spool via special purpose clamps.

Separate choke and kill lines were installed on the subsea snubbing BOP stack, since it was anticipated that extensive circulation would be required after the drill pipe was tied back. The choke and kill lines allowed circulation without sustained pressure on the riser spool. Dual valves (one HCR and one fail-safe) were installed on the subsea drilling cross for safety. Choke and kill lines were fabricated of 5-in. drill pipe and connected to the riser spool via special purpose clamps.

Equipment and services. Since snubbing operations are rarely undertaken in the region, virtually all snubbing and auxiliary equipment had to be mobilized from the U.S. for the project. To minimize planning difficulties, ONGC requested that as much equipment and services as possible be supplied via a single source. A detailed equipment and services list, along with a scope of work, were developed and sent to three snubbing companies in the form of a request for bids to act as general contractor for the project. The equipment and services to be supplied by the project general contractor included:

• Snubbing unit (400K minimum) and related support equipment
• BOP/pressure control equipment and specialist personnel
• High-pressure riser equipment and specialist personnel
• Fishing tools and specialist personnel
• Tongs and tong operator
• Slickline equipment and operator
• Wireline set bridge plugs for all pipe strings.

ONGC supplemented personnel and equipment requirements by supplying:

• Rig and rig support duties (catering, crew change, etc.)
• Electric wireline services
• Pumping personnel and equipment
• Cementing and drilling fluids personnel, material and equipment
• Communications
• Diving support personnel
• Marine vessels.

Other services, such as non-destructive testing (NDT), specialty fabrication, machine work, etc., were also obtained locally.

Mobilization planning. The temporary import of the massive amount of equipment was a monumental logistical challenge. The process was divided into two distinct phases – loading and airfreight to India, and offloading, customs clearance and inland transportation to the wellsite.

The first phase was arranged by the general contractor. Transportation involved the airfreight shipment of over 496,000 lb of equipment, plus the timely arrival of specialized personnel. Arrival of the equipment and personnel was coordinated with the end of the monsoon season and the mobilization of the jackup.

Customs clearance and inland equipment transportation was performed without incident and all equipment arrived at the rig per the mobilization plan.

PROJECT IMPLEMENTATION

The operation began with a thorough inspection of the existing subsea BOP stack. The inspection was performed and no leaks or other anomalies were found.

Rig-up. The first step in installing the snubbing equipment and high-pressure riser was to remove the upper annular from the existing 18-3/4-in. subsea BOP stack. This annular had a maximum working pressure of 5,000 psi and had to be removed to provide access to the 10,000-psi-rated Cameron (CIW) 27 hub connection below it. The following procedure was used to remove the annular BOP:

• Pick up one joint of 5-in. drill pipe pin end first.
• Install inverted 5-in. drill pipe elevators onto box end of drill pipe joint.
• Attach slings to eyes of elevators.
• Lower the first joint of 5-in. drill pipe (upside down).
• Install 4-1/2-in. IF box to box crossover.
• Run in hole with 5-in. drill pipe to top of annular BOP (as confirmed by diver).
• Position rig cantilever exactly over annular and attach slings to BOP.
• Pull 10,000 lb.
• Install two, 10-ton come-a-longs from BOP bonnets to BOP frame.
• Remove top ring from annular (remove with rig air hoist).
• Disconnect CIW 27 hub connection.
• Lift annular BOP clear of subsea stack and raise to water surface.
• Position dynamically positioned dive vessel under annular.
• Set annular onto deck of dive vessel and disconnect lift slings.

The annular BOP was removed according to plan, although some problems were encountered accessing the shuttle valve assembly on the annular as it was positioned behind the kill line stab lock ring. This required removal of the upper portion of the kill line and the kill line hydraulic stab.

Installing HP riser and BOP. Before installing the high-pressure riser, the 5-in. drill pipe sections that would serve as the choke and kill lines were run in a conventional manner through the rotary. Once required lengths were in place, they were suspended from the main deck using cables and pad eyes that had been installed for that purpose. Each section of drill pipe had 3-in. steel hose connected to the lower end. These hoses would connect the choke and kill lines to the drilling cross once it was in place.

The subsea BOP stack was first assembled in three sections on the deck of the jackup. The bottom section of riser was picked up and lowered through the rotary, and the drilling package was skidded over the main deck so the lower riser connection could be made up onto the existing section of the BOP stack. This section of the BOP stack accompanied the riser section so that the drilling package could be positioned over the middle BOP section. The flange connection was made up and the process was repeated for the third (lower) BOP section. The drilling package was then repositioned and the entire assembly was lowered. Once the first riser section was run, subsequent riser sections were picked up from the deck and installed in a conventional manner using a special spider assembly and elevators.

Divers confirmed that the bottom connection of the riser/BOP stack was approaching the upper connection of the existing subsea BOP stack. Helmet-mounted and hand-held subsea cameras were used along with additional audio/ video units on the rig floor to monitor the lowering of the riser assembly until the two halves of the CIW 27 clamp were mated. Divers then tightened the CIW 27 clamp with hydraulic wrenches, connected choke and kill lines to the drilling cross and attached the 5-in. drill pipe lines to the riser sections. The riser was tensioned to 250,000 lb. (about 50,000 lb over-pull) and this load was then transferred to the spider assembly on the rig floor.

Snubbing unit and surface BOP. With the riser installed, the snubbing unit and surface BOP equipment was rigged up in a conventional manner. Fig. 4 shows the snubbing BOP arrangement.

Testing. A detailed testing procedure was developed, and all pressure control components (with one exception) were tested to 250 psi and 9,000 psi. Pressure tests were performed by closing the inverted blind rams on the bottom of the 11-in. subsea BOP assembly. These rams were placed in this position specifically for testing purposes.

The one exception was the CIW 27 clamp connection where the additional (11-in.) subsea BOPs were connected to the existing (18-3/4-in.) subsea BOPs (i.e., below the inverted blind rams). This connection could only be tested to 5,000 psi, at which point the SBRs would leak due to the pressure differential applied from the top. Water-soluble oil (which has a very visible white color in water) was used to test this connection, while subsea cameras on the BOP stack were used to ensure that the leak was through the ram and not from the connection itself.

WELL CONTROL INTERVENTION

Intervention was initiated by opening the 18-3/4-in. SBRs. Drill pipe pressure was 4,250 psi. Attempts were made to lower this pressure by lubricating mud, but this did not yield significant results.

The top of the sheared drill pipe had to be milled to allow access by the high pressure pack-off overshot that could be used to tie back the drillstring. The 18-3/4-in. variable bore rams (VBR) were opened so that milling could be done on the top portion of the sheared drill pipe. Milling was started using 5-in. drill pipe with 4,250-psi surface pressure. Leakage occurred during the milling process (stripper ram leakage), which caused surface pressures to increase dramatically. However, these ram leaks did not pose an unmanageable safety threat. Surface pressure increased to 6,500 psi following several complete ram seal failures. Only 6 in. were milled off the top of the drill pipe before surface pressure rose to an unacceptable level. An additional 8 in. would have to be milled to allow the high-pressure (10,000 psi) pack-off overshot to properly “swallow” the drill pipe.

Fig. 5. After several failed attempts to set a bridge plug using wireline, it was decided to install a second, smaller snubbing unit onto the drill pipe. This 150K unit used 1-1/4-in. macaroni tubing to clean out the drill pipe/ tubing string to a depth of 9,600 ft.

The milling work created an ID at the top of the sheared drill pipe that was at least as big as the ID of the 5-in. drill pipe tool joint. Thus, a decision was made to set a wireline bridge plug in the bottom of the tubing string. This would allow a reduction in the surface pressure, assuming the tubing string was intact (as expected from earlier operations).

Several attempts were made to run in the hole with electric line. Each attempt was unsuccessful due to the condition of the mud (which had not been circulated for several months) and the extreme surface pressure. After a thorough evaluation of the available options, it was decided that the best course of action would be to install a second, smaller snubbing unit onto the drill pipe. This 150K unit would use macaroni (1-1/4-in.) tubing to clean out the drill pipe/ tubing string.

Fortunately, this contingency had been discussed during the project planning, and a small snubbing unit, along with necessary BOP equipment, tubing string and handling tools, was mobilized from the U.S. in a remarkably short time. Operations were temporarily suspended while the additional equipment was transported.

Once at the wellsite, the small snubbing unit was rigged up, tested and running pipe in less than 48 hr, Fig. 5. The 1-1/4-in. tubing was used to clean out the combination drill pipe/ tubing string to a depth of 9,600 ft. Seawater was used to displace the old drilling mud in the drill pipe/ tubing string. Once the pipe string was circulated clean, seawater was displaced with 11.6-ppg CaCl 2 . This reduced surface pressure to 2,500 psi. The 1-1/4-in. tubing was removed and the 150K snubbing equipment was rigged down.

A second attempt was made to set the bridge plug with wireline. This time, surface pressure was only 2,500 psi and there was clean brine in the pipe string. The bridge plug was set in the bottom joint of the 3-1/2-in. tubing and surface pressure was bled to zero in 500-psi increments.

All surface pressure was now contained below the 18-3/4-in. subsea pipe rams. This allowed the milling to be continued with no surface pressure. An additional 8 in. was milled from the top of the sheared drill pipe. This allowed the high-pressure pack-off overshot to be latched onto the 5-in. drill pipe and tested.

With the pack-off overshot in place, the drill pipe and annulus were isolated. Communication could be reestablished once a flow path was developed. The drill pipe was tensioned to the previous string weight and the 18-3/4-in. subsea pipe rams were opened. Initial annulus pressure was recorded at 3,200 psi.

As expected from earlier operations, the drill pipe/ tubing string was stuck due to bridging in the annulus. While attempting to get a preliminary idea of the stuck point via stretch measurements, the 3-1/3-in. tubing string parted at 1,657 ft. Since annular bridging was apparent and deep (based on preliminary stretch data), it was decided to attempt bleeding the surface pressure. The surface pressure was bled to zero in 250-psi increments.

It was decided that the safest and most economical solution would be to abandon the B-24-2 well with cement plugs. The 3-1/2-in. tubing fish was removed and the tubing was milled off and re-latched. Circulation was established by perforating the 3-1/2-in. tubing at 6,004 ft. The well was circulated at this depth with 14.0 ppg mud. A significant amount of gas was removed from the annulus, but the well remained static after circulating.

The first abandonment plug was set 656 ft inside the tubing and in the tubing/ casing annulus from 6,004 ft to 5,348 ft. The plug was subsequently tested to 3,500 psi after being allowed to set.

A second balanced cement plug was set from 1,549 to 1,220 ft and similarly tested. Finally, a wireline-set bridge plug was set in the 9-5/8-in. casing at 1,148 ft.

CONCLUSIONS

The B-24-2 well represented a massive technical and logistical challenge. The rare circumstances of the situation required a unique combination of technical expertise to develop and implement the safest and most economical solution. Personnel from ONGC and third-party specialists had to rely on their extensive experience in order to adapt existing technology to an unusual and potentially dangerous situation.

The possibility of catastrophic failure existed on several levels during the entire well control project. The consequences of such a failure remained paramount throughout the planning and implementation of the solution. The authors believe that a critical, complex project such as the B-24-2 intervention can only be performed safely and successfully when there is an exceptionally cooperative well control and engineering effort on the part of the operator and specialist contractors involved. 

THE AUTHORS
	David Barnett, engineering manager for Wild Well Control, Inc., has over 20 years of drilling, snubbing, coiled tubing and well control experience. He has been involved with the planning and implementation of numerous relief well and high pressure snubbing and well recovery operations. He also has been instrumental in the development of blowout contingency plans (BCP) for a large number of operators ranging from small independents to multi-national corporations. In addition, he is involved in all aspects of engineering division projects including platform design, dynamic well-kill modeling, training and research into improved well control equipment and techniques. Mr. Barnett earned a BS in mechanical engineering from the University of Houston, and he is a member of SPE, API, AADE, ASME and IADC.
	D. C. Tyagi, a mechanical engineer, has worked in India's Oil & Natural Gas Corp. (ONGC) for the last 19 years. He is a member of the Crises Management Team of ONGC and has been associated with almost all difficult well situations in ONGC. He is currently working as a chief engineer in the Mumbai Regional Business Centre and looks after operations of the drillship Sagar Bhushan.
	A. K. Mehra, a mechanical engineer, joined ONGC in 1969 and since then has been working on various assignments including supervising drilling and workover operations in deep wells. He is credited with many pioneering projects such as upgrading the drillship Sagar Vijay for deep water drilling. He was associated with well control operations on the subsea well, B-24-2, and is currently general manager (Drilling) and head of the Drilling Business Group in the Mumbai Region of ONGC where 20 offshore rigs are operating. Mr. Mehra's ME degree is from Punjab University, and he also holds a law degree and an MBA. He has authored many R&D articles.