Part 11-Relief wells: Advancements in technology and application engineering
make the relief well a more practical blowout control option
John W. Wright, Blowout Advisor, John Wright Company,
Houston,
and
L. Flak, former Wright, Boots & Coots employee.
This article describes evolution of the relief well in blowout control in terms of
technology, strategy, planning and use . Concepts reviewed explain when a relief well
is appropriate and the final section describes basic steps in design and execution of a
relief well operation.
More blowouts now warrant evaluation of a relief well as a primary control option. In the past, relief well developments in technology and strategy occurred only during unique blowout control operations. Now, how ever, a continuous process of improvement has evolved, aided by hydraulic kill models. As a result, strategic planning and new technology provide a more adaptive and efficient control option.
The relief well has traditionally been a last resort when other surface
kill efforts fail. This has changed with increasing technology requirements for
horizontal, deep, offshore, hostile environment, or high pressure wells. Questions arose
whether blowouts of some wells could be killed at all, especially with the possibility of
under ground blowouts. Fortunately, relief well advancements paralleled this period of
technology growth and now provide viable blowout control options. The operator of a
blowing well will likely consider surface capping methods before snubbing or relief well
options. Some of the events influencing choice include:
If a well clearly cannot be capped, the decision is simple-drill a relief well. But if it is uncertain whether the well once capped can be killed, then additional options remain. These include (1) rig up a snubbing or coil tubing unit to run a kill string and perform a circulation kill, (2) drill a relief well, or both. A planning team must quickly evaluate each option, associated safety risks, pollution, escalating severity, logistical obstacles, public concern, available resources, and other factors that might override preferred strategy. Complex, informed decisions must be made, especially when considering parallel surface and relief well operations.
To make a decision, the operator must be aware of changes in technology, applications, planning technique, and demonstrated success. Starting a relief well plan is not only cheap insurance should initial strategy fail, but a demonstrated means of efficiently killing blowouts.
A change in relief well technology has always prompted a refinement in application strategy, often broadening the range of uses. Strategy has become a tool in itself, supported by software models and complex decision analysis.
The original purpose of a relief well was to relieve pressure on a blowing formation by drilling a vertical well around the blowout and producing it (them) at high rates.
A directionally drilled relief well on a prolific cratered blowout near Conroe, Texas, in 1933 marked the first mile stone in relief well development. New borehole survey instruments and the openhole whipstock allowed controlled directional drilling and first changes in relief well procedure from drilling vertical producing wells. A directionally drilled relief well intersected the flowing reservoir below the blowout surface location, water was pumped into the reservoir, and the well was successfully killed.
Between 1933 and 1970, most relief wells followed this basic strategy. Due to limitations of borehole surveying instruments, actual distance between blowout and relief holes was subject to significant bottomhole position uncertainty. Two or more relief wells thus were often used to increase probability of gaining hydraulic communication.
Early procedure was to pump water into the reservoir to communicate with the blowout well, sometimes killing the well by flooding the reservoir. Other cases required pumping weighted mud into the blowout well bore after gaining communication through a channel created by water flow, acid treatment, or fracturing. However, increased drilling depths, high GOR producers and low permeability reservoirs resulted in blowouts that could not be killed with such methods.
In 1970, Shell Oil Co.'s Cox 1, a 22,000-ft Smackover exploratory well, blew out near Piney Woods, Mississippi. Accurately drilling a relief well to that depth with existing techniques was doubtful. This challenge led to the first direct intersection of a blowout tubular using a detection method. Wireline instruments were developed to detect proximity of a tubular by measuring distance and direction from the relief well to the blowout casing. Ultimately, the well was intersected and killed at 10,500 ft, with communication gained by perforating from the relief well to the blowout. This success was the beginning of the modern relief well, establishing strategy and planning for future relief well projects and the basis for commercial casing detection instruments (outside the Soviet bloc).
Magnetostatics. In 1975, a Gulf of Mexico blowout provided economic incentive for developing the first commercial magnetostatic or "passive" casing detection tools, still in use. These instruments measure perturbations of the earth's magnetic field caused by remnant magnetic poles on the blowout casing or drillstring. Basic physics allows determination of relative distance and direction from relief well to blowout. The outer range of detection is 20 to 40+ ft depending on magnetic pole strength.
This method was used on most relief wells through the mid-1980s requiring ranging on tubulars in the blowout well.
Specialty kill fluids. In 1976, specialty fluids were used during a unique cratered blowout in the Persian Gulf from a high permeability gas section of the Asmari formation at 3,500 ft. Hole size was 17-1/2 in. With casing at 1,100 ft. Four relief wells gained hydraulic communication with the borehole, but were unable to control the flow with conventional kill fluids. This resulted in first use of polymer systems as kill fluids. Two polymer types were pumped through separate relief wells. An extremely viscous, cross-linked guar gum was pumped into salt cavities above the reservoir, and a high molecular weight HEC polymer was pumped into the reservoir matrix. The guar filled the cavities and decreased gas flow while the HEC blocked off loss of kill fluid to the vuggy, fractured reservoir matrix.
A similar technique was successful on Mexico's offshore Ixtoc blowout in 1980.
Dynamic kill. In 1978, Mobil Oil documented the technique of "dynamic kill" on a prolific gas blowout in Arun field, Indonesia.4 The technique involves circulating a light initial fluid, such as water, with sufficient friction pressure to kill the blowout (hence the name "dynamic"), followed by mud with sufficient density to contain reservoir pressure. Advantages include its use when kill pressures in the well bore must be developed in a controlled manner to prevent formation fracture; simple hydraulic calculations; and use of the relief well drillstring for real time measurement of BHP during pumping. Disadvantages include high hp requirements for killing a well with a light fluid. This technique laid the foundation for future engineered kill procedure designs.
Electromagnetic casing detection. In 1980, another blowout in the Gulf of Mexico led to commercial development of the first electromagnetic "active" ranging instrument. By applying AC electric current to the blowout tubulars, a wireline instrument in the relief well can detect the induced AC magnetic field. Instrument sensors measure field direction and intensity, providing data for calculating relative distance.
A blowout in 1982, using a modified technique with downhole current injection, demonstrated that casing could be detected at a range of at least 200 ft. The technique efficiently located blowout tubulars for a direct intersection. Casing detection and other developments in surveying and MWD proved a technique for triangulating the blowing well, reducing plugging and sidetracks. This again changed basic strategy for designing relief well trajectories. Accuracy meant savings; few blowouts after this date involved two relief wells.
Borehole surveying technology and procedures began to advance in the 1980s. Small diameter north-seeking rate-gyro systems, with greatly increased cased hole survey accuracy, became commercially available in 1982. MWD technology advanced rapidly in this period, particularly with respect to reliability, transmission speed, smaller sizes and directional instrument accuracy. Major advances were also made in under standing and correcting error sources inherent in MWD and electronic multishot surveying with respect to sensors, mechanical flexures in the BHA and BHA magnetization.
Due to a better understanding of the earth's magnetic field and the ability to sample raw data from accelerometer and magnetometer arrays, quality control of surveys became easier. Relief wells could be targeted more precisely with better information and more confidence.
Surface kill equipment. Basic requirements have not changed, but advances in high pressure hydraulic fracturing in the 1980s increased efficiency of controlling a kill operation. Skid-mounted portable high pressure pumping assemblies are easily manifolded and tandem-stacked if necessary for offshore applications. Higher hhp frac pumps are available for many land operations. All pumps can now be remotely operated from a single control point, facilitating quick changes in pump rates or kill fluids, and making kill pumping operations more coordinated and controlled. Modular high pressure manifolding systems, once custom manufactured, are now avail able in various sizes for up to 20,000 psi discharge pressures.
Blowouts offshore pose problems in accessing wellheads, pumping equipment and kill lines. Steel flex hose technology allows connections to sub sea wellheads in deep water. Remotely operated valves, relief valves and hydraulic disconnect units make quick isolation of a kill line safe and practical during the kill.
Computer vans or offshore modules are available for monitoring pressures and flowrates required during a kill in a single air-conditioned quiet room. The kill supervisor can direct operations in a non-hostile environment with multiple television screens plotting various data sets for quick analysis. Self-contained stimulation vessels are now available in many offshore areas of the world and make suitable kill platforms in some situations.
Steerable systems. In 1988, fully steerable directional drilling systems were first used on a relief well. Using stabilized bent housing motors in various configurations with a reliable and accurate MWD system facilitates precision directional work required to drill complex relief well trajectories for ranging triangulation and direct intersections. Relef well strategy changes made this same year combined relief well trajectory and electromagnetic ranging constraints into better planning for more accurate and efficient placement of the relief well.
The result. In 1989, the result of 20 years of new technology and strategy proved itself in the North Sea on the Saga petroleum 2/4-14 blowout, with a direct intersection of an 8 1/2-in. borehole at a depth of nearly 5 km.(7) No sidetracks were requred and only nine electromagnetic fixes were made.
The project further led to development of a sophisticated, fully dynamic, two-phase hydraulic kill simulation. This software allows complex evaluation of many kill scenarios with various sensititivites, to determine the most efficient kill method. This capability has become a powerful tool in optimizing relief well kill strategy.
Blowout contingency plans were instituted for many international operators int he early 1990s. These plans document general emergency procedures, apply newly developed control strategies to specific wells and offshore structures, and provide a basis of examination. The process has helped define and address critical problems that might be encountered in controlling a blowout with a relief well. It also allows for continuous strategy refinement.
Horizontal drilling activity has resulted in BHAs capable of producing controllable dog-leg severity rates >20 deg/100 ft, increasing options for relief well trajectory design, particularly for shallow blowouts. Rugged rate-gyro survey sensors provide instrumentation for steering tools used while drilling with a mud motor. This is especially useful for drilling a relief well next to casing in a vertical blowout. Recent advances in borehole survey technology provide small diameter full inertial navigation systems, using both steel and laser gyros capable of mapping borehole trajectory with an uncertainty approaching 1 ft/1,000 ft of hole depth in a fraction of the usual gyro survey time.
Electromagnetic detection advances have reduced uncertainty in relative distance measurements by using better measurements of the electromagnetic field.
Direct measurement of distance now is possible independent of the amount of current flowing in the target at distances up to 30 ft with uncertainties of +5% of the distance. Another tool, providing measurement along the z-axis, enables placement of a vertical relief well over a horizontal blowout or other high approach angle situation. Where surface access of the blowout is possible, such as a simultaneous snubbing operation, an AC electromagnetic source can be deployed by wireline in the blowout well. A sensor in the relief well measures the induced magnetic field and determines distance and relative direction with uncertainties less than +10% of the distance. These casing detection options support a broader range of relief well design possibilities.
Specialty fluids today. Two blowouts in 1993 resulted in further refinement in application of specialty kill fluids when conventional fluids (water, brine, weight mud) did not work. Conditions that call for a special fluid, or a two part reactive fluid mixture, that can set up quickly and plug off and/or separate two flowing zones include:
Fluid types used include crosslinked and linear polymers, with gel time and strength controlled by temperature and pH. Other two-fluid reactive mixtures must be pumped separately or in slugs similar to pumping a gunk plug for lost circulation. Soft plugs (diesel oil, bentonite and cement that react with water) can be successfully used in specific situations. Chemicals that will form hard plugs when properly mixed can seal off a borehole or aid in the killing process when combined with heavy mud and/or cement. Such a plug controlled a prolific gas blowout in Argentina in 1993.
This guide outlines non-routine steps in planning a relief well based on proven technology. The design usually requires several revisions before an acceptable plan is achieved (Fig. 33). The blowout scenario controls the planning process. If the scenario or severity changes, the relief well plan may require dramatic changes.
When to start a relief well. The first step is to determine if the blowout well can be capped and killed by bullheading or circulating down existing tubulars. If there is significant uncertainty in the ability to cap the well, a pre-emptive relief well planning team should be formed (within 48 hrs of the blowout). Cost of a plan is cheap insurance should a relief well be needed and the relief well can often serve as a replacement well.
If the well cannot be capped, the relief well(s) can then start as soon as possible. If the well can be capped but not killed, then use a snubbing or coil tubing unit for a circulating kill, or drill a relief well, or start both operations simultaneously.
Planners complete basic snubbing and relief well evaluations and identify weaknesses. Blowout snubbing operations that require fishing or have high gas flowrates have a high incidence of problems. If there is major uncertainty in snubbing success, then a relief well should be started parallel with snubbing. Advantages of either may be altered by overriding factors such as pollution, safety, etc.
It may be more cost effective to drill a relief well rather than snub if:
Task force. Organize a dedicated task force for planning and executing the relief well. Depending on project size, this may be a few people or a large organization. The leader should be a senior drilling engineer or drilling manager from the operating company. There may be an office planning team and a field execution team. Teams are normally broken into two functions, one planning the kill operation and the other planning the drilling and intersection pro gram. At least one relief well advisor and one senior drilling engineer should be assigned to each group. Support personnel are added as needed depending on project size.
Initial decisions. Once the task force is formed, it should at least consider these questions:
There should always be a mechanism to update and review the decision-making process, and a careful pacing of the decision itself. Though there is always a sense of urgency in evaluating decisions during a blowout (and many can be made quickly with the help of a relief well advisor), some decisions take time and research to clarify
Planning process. Relief well planning is a repetitive and parallel process requiring simultaneous evaluation of the kill pumping program and the drilling and intersection program. The kill point governs outcome of both programs. The kill point must be weighed against what is best for killing the well and what is best for drilling and intersecting. Experience will help make initial assumptions. The kill team then analyzes the kill program based on the given point and the drilling team analyzes the drilling and intersection pro gram based on the given point. If either group finds the chosen point unacceptable, it will be moved up or down the well and re-analyzed until an acceptable plan is reached.
Once the chosen relief well design is presented, it should be continuously analyzed for safety, logistics, probability of success and economics (see box). Coming next: The concluding article in this series will discuss the developing role and interrelation between oil well firefighting companies, blowout engineering advisors, service companies and operators to better manage blowout control hazards.
| Table: Relief well planning considerations | |
|---|---|
Iterative systematic planning
|
Hydraulic design
|
Surface equipment
|
Kill procedures
|
Establishing kill point
|
Borehole uncertainty
|
Initial search depth
|
Surface location
|
Relief-well geometry
|
Establishing fluid communication
|
Well control incident management and critical alliances. The final ariticle in this
series on blowout control, summarizing the contents of the series and providing insight as
to the future of blowout control.
Next Article
Literature Cited
1. Bruist, E. H., "A new approach in relief well drilling,"
SPE 3511, New Orleans, La., 1971.
2. Morris, F J., R. L. Walters and J. P Costa, "A new method of
determining range and direction from a relief well to a blowout" SPE 6781, Denver,
Colo., 1977.
3. Arnwine, L. C. and J. W. Ely, "Polymer use in blowout
control," SPE 6835 Denver, Colo , 1977.
4. Blount, E. M., and E. Soeiinah, "Dynamic kill: Controlling wild
wells a new way " World Oil, October 1981.
5. West, C L., and Kuckes, A. F., "Successful ELREC logging for
casing proximity in an offshore Louisiana blowout," SPE 11996, San Francisco, Calif
,1983.
6. Flak, L. H. and W C. Jr Goins, "New relief well technolngy is
improving blowout control" World Oil, Dec. 1983 and January 1984.
7. Leraand, F., J. Wright, M. Zachary and B. Thompson, "Relief well
planning and drilling for a North Sea underground blowout," JPT, March 1992.
The authors
John Wright's photo and biography appeared in Part 1 of this series. Please see World Oil, Nov. 1993, page 78.
L. Flak is a former John Wright Company employee.
