Jack M. Bailey
Chevron U.S.A. Production Co.
Rangely, Colo.
A well control problem during a routine workover, in which a packer unseated and tubing came out of the well, has led to new operational practices and a renewed commitment to necessary precautions.
On Monday, Apr. 11, 1994, a well control event occurred that caused 2 7/8-in. tubing to be ejected from the Fee 83X well in the Rangely Weber sand unit in Colorado. The tubing landed in a wash next to the well. The well was contained and secured without personnel injury or damage to the rig, the well, or the location.
This well control event on Fee 83X was a very unpleasant experience, and many people can second guess and speculate on causes, actions, and mistakes made. Nevertheless, all personnel involved did their best and performed their duties well. Of importance is that several lessons were learned on improving well work operations.
BACKGROUND
The rig had been on location for 17 days performing a workover, which included fishing out the selective injection equipment, cleaning out the well bore, adding perforations, and performing an acid stimulation job. When the well control event occurred, new injection equipment was being installed, and the well was being prepared for water/CO2 injection.
During the course of the job, the well was controlled with 11.2-ppg CaCl2 fluid. To keep bottom hole pressures down, the well was flowed back each night to a collection station. This practice is standard for working on high pressure CO2 wells; otherwise, the maximum density (11.6 ppg) CaCl2 fluid available would not control the wells the following morning.
Prior to the incident, a Baker Oil Tools Lok-Set packer with 23 joints of fiber glass tailpipe was run in and set above the perforations at 5,590 ft.
The assembly had a 6 ft tubing sub and a Baker on/off profile above the packer with a Baker FWG plug installed in the on/off profile to provide a mechanical well bore seal to contain the formation pressure. This completion is standard for injection wells in the field and provides a mechanical well control device to contain pressures while the work string is laid down and the cement-lined injection tubing and isolation packer are run in. Position 1 in Fig. 1 (21779 bytes) shows the completion equipment in the well bore prior to the incident.
TUBING EJECTION
- After the Lok-Set packer was set, the tubing was disconnected from the on/off tool and the CaCl2 fluid was displaced out of the hole with source water.
- The tubing was reconnected to the on/off tool, and the casing was pressure tested.
- An acid stimulation job was pumped and then flowed back for clean up. The well was then flowed to the collection station overnight.
- On wire line, a Baker FWG plug was set in the on/off profile. The tubing was disconnected from the on/off tool and then pulled out of hole and laid down.
- An on/off skirt and five joints of 2 7/8-in. injection tubing were run in the hole, and then the well started to flow. The well was shut in. The well was then opened to try to bleed off trapped CO2 pressure, but with no success. The well was lubricated with kill fluid and bled until it was dead.
- The leak was suspected to come from the packer or FWG plug, so the five joints of injection tubing were laid down.
- The on/off skirt was picked up and run in the hole on 37 joints of 2 7/8-in. work string tubing. The well started to flow again, so the well was shut in (Position 2, Fig. 1) (21779 bytes). CO2 gas was then bled through the flow line, and then the tubing moved out of the hole 10 ft (Position 3, Fig. 1) (21779 bytes). The pressure on the annulus and tubing bled to zero. It was assumed that gas was out of the well bore and the packer and FWG plug were containing formation pressure (as learned later, this assumption was incorrect).
- The blowout preventer (BOP) was opened, and the tubing was moved to continue running it in the hole. The tubing was then ejected from the well bore like a rocket, and all the on-site personnel ran to safety. The floor hand actuated the accumulator to close the BOP as he ran from the rig. The pipe rams closed on the tubing and stopped the packer from exiting the well bore (Position 4, Fig. 1) (21779 bytes). The annular preventer was then closed on the tubing to contain the well.
All personnel were located and accounted for at the staging area, and there were no injuries.
A total of 37 joints of tubing were ejected from the well bore in less than 20 sec. The wash, or dry stream bed, where the tubing landed was on the opposite side of the rig from the doghouse. There were no injuries because all the personnel ran toward the dog house away from the ejecting tubing except one rig hand who stayed under the substructure.
The top tubing collar lodged in the rig elevators, causing the tubing to bend and exit through the tubing tongs, over the rig floor and into the wash.
ASSESSMENT
After all had settled down, the cause of the problem was assessed. During the bleed down of the second kick (Step 7 above), the downhole packer had released and traveled up the well some 4,500 ft. It ran into the bottom of the on/off skirt attached to the 37 joints of the work string. When the packer hit, it preset slightly and sealed the well bore. Thus, the pressure was able to bleed to zero (during Step 7).
At the time, there was no idea that the packer had moved up hole. When the BOPs were opened and the tubing was moved, the packer released again and acted as a giant piston, pushing the 37 joints of tubing out of the well.
In hindsight, the procedure should have been to try to pump down the well when the tubing moved out of the well 10 ft. Maybe the packer would have released, allowing kill fluids to be pumped into the well, indicating the packer or FWG plug had failed. The well could have then been killed by bull heading and controlled for subsequent well work operations.
REMEDIAL ACTION
After the initial assessment of the cause of the problem, fluid was pumped down the annulus and the tubing with no success. The on/off skirt had been caught and stopped by the pipe rams, and the packer was packed off again. The well was left shut in for the night.
The rig was cleared after several hours of cutting and untangling tubing. The top joint of tubing was bent at a 45 angle over the rig floor. The efforts then focused on pulling the FWG plug so kill fluids could be pumped down the well:
- Through a series of pulling and pushing with the winch line and boom trucks, the top joint was reasonably straightened.
- The wire line lubricator was installed on the top joint of tubing. An attempt to pull the plug resulted in stuck wire line tools that, had latched the plug, Because of the bends and kinks in the top joint, the spang jars could not be used to release the plug.
- The top joint of tubing was cut 1 ft above the rig floor and lifted up along with the lubricator using the winch line to ensure that the string of wire line tools did not come up with the tubing.
- The wire line tools were broken apart at the nearest connection. A 2 1/2-in. thread was cut on the top joint stub. A tubing sub was then connected, and a safety valve was installed.
- The wire line lubricator was made up and pressure tested against the FWG plug, which also tested the newly made connection.
- The FWG plug was pulled successfully. The well was then flowed back overnight.
- Kill fluid was pumped (bull headed), successfully killing the well.
- The annular BOP was opened, but the pipe rams would not open because they were damaged by the impact of the on/off tool. While kill fluid was pumped down the well, the BOP doors were opened, and the damaged rams were removed. New rams were installed, and the packer was retrieved from the well. The packer looked new except for swelled packing elements caused by CO2 exposure.
- The BOPs were tested, and subsequent well work operations continued.
PACKER EVALUATION
The Lok-Set packer is believed to have released because of one or more of these possible reasons:
- The packer was set in deteriorated casing.
- When set, the packer was not fully engaged because of soft setting weights.
- The packer slips could not set into the casing because of hard scale not removed in previous well bore cleaning efforts.
RESULTS
During the course of this well control event and the subsequent analysis, many important lessons were learned. A round-table discussion, which included all toolpushers and workover representatives, resulted in the following procedures and considerations for future well work operations:
- The setting procedures for the Lok-Set packer were revised: Set down full string weight, and pull 30,000 lb over string weight. This greater pull will ensure a harder set so the packer is anchored to the casing better.
- Precautions should be taken to ensure that the packer seats are clean and free from scale buildup by making a scraper or string mill run sometime during the workover before the packer is run.
- Whenever possible, kill fluid should be kept in place above mechanical sealing devices until it is time to circulate packer fluid in place. Also, the hole should be kept full, or fluid should be pumped into the hole at all times.
- Casing caliper logs should be run in wells where corroded or bad casing is suspected. Packer setting depths should be chosen from the log, and the depths should be double checked to ensure they meet Colorado Oil & Gas Conservation Commission (or other regulatory agency) regulations on packer setting depths.
- After a packer and plug, which will be used for a well isolation shutoff device, are set in place, they should be pressure tested to ensure integrity.
- The coordination of shutting in offset wells before and during workovers should be reemphasized with district teams.
- Well control training should continue, and the techniques and methods learned should be practiced. All personnel should continually be on the lookout for out-of-the-ordinary events that can lead to a well control situation.
- The contingency plans for CO2 events, such as injection line rupture, well work, and any other applicable operations, should be revisited.
Copyright 1995 Oil & Gas Journal. All Rights Reserved.