IDENTIFYING CRATER POTENTIAL IMPROVES SHALLOW GAS KICK CONTROL

Luiz Alberto Rocha
Louisiana State University/Petrobras
Baton Rouge

Adam T. Bourgoyne
Louisiana State University
Baton Rouge

An understanding of sea floor crater mechanisms can help drillers determine whether to divert or shut in a shallow gas kick.

Proper well planning that considers potential shallow gas sources can eliminate some of the more common failure situations with both diverting and shutting in of a well.

Current well control practice for land and bottom-supported marine rigs usually calls for shutting in the well when a kick is detected, providing sufficient casing has been set to keep any flow underground. The casing and surface equipment must have an adequately high working pressure to ensure that formation fracture occurs before the equipment fails.

Even for high shut-in pressures, an underground blowout is preferable to a surface blowout.

An operator may choose to divert the flow if the surface casing is not set deep enough to keep the underground flow outside the casing from breaking through the sediments to the surface. Once the flow reaches the surface, a crater may form at the sea bed, possibly sinking or damaging the rr. Craters increase the difficulty and time required to kill a blowout.

SHALLOW GAS

In some marine environments where abnormal formation pressures may be encountered at very shallow depths, conventional blowout prevention equipment and procedures may be of little benefit. Well control operations could quickly become difficult if permeable, gas-bearing formations are drilled.

Many shallow gas accumulations are slightly overpressured in the upper portion of the reservoir because of the density difference between the gas and the surrounding water.

Abnormal formation pore pressures that approach the formation fracture pressure may occur in sand lenses because of gas migration from below and along fault planes.

Onshore, the presence of unconsolidated sediments over shallow artisan water reservoirs can lead to severe structural damage if a blowout occurs. The consequences of a blowout in such an environment include the following:

Charging of the shallow water reservoirs and subsequent failure of the rock by hydraulic or shear fracture
Severe borehole erosion followed by a pressure drop in the well, upward flow of gas or water, and caving of the surrounding formations.

To control a shallow kick (mainly gas), the operator must decide whether to divert the flow or to shut in the well. Additionally, well control failure can allow the formation fluid to migrate up, reaching shallow, poorly cemented or unconsolidated sediments and forming a crater.

DIVERT OR SHUT IN?

If a kick occurs with the casing string set at a shallow depth where formations with low fracture gradients prohibit holding back pressure, the operator will have to decide quickly whether to divert the flow or shut the well in.

On a bottom-supported rig, controlling a typical shallow -as kick using a diverter system is summarized as follows:

The gas kick is detected after some amount of influx.
The diverter system is actuated. The vent line is opened, and the annular blowout preventer is closed, both automatically.
If the well plan calls for a dynamic kill operation, mud is pumped at the maximum flow rate to attempt to regain control.

Diverted kicks tend to produce large amounts of abrasive solids at high velocity, which often destroys surface equipment. Recent U.S. Minerals Management Service statistics indicate that diverter systems have failed during approximately 46% of well control operations for gas flow. The six most common modes of well control failure using diverter systems are a damaged line, a failed valve, a plugged line, a failed annular preventer, an ignited flow, and a broached sea floor.

The advantages of shutting in the well include the following:

Further kick development is stopped immediately, and the kick is held to a minimum value.
The formation pressure can be monitored, and kill weight mud can be determined.
Conventional well control methods can be used to circulate out the kick to kill the well.
The well is essentially under control.

If the well is shut-in, the following consequences are possible:

Shut-in drill pipe and casing pressure can stabilize at safe values without fracturing the casino seat (best case).
The casing can fracture during shut-in, causing an underground blowout that stays confined in the subsurface.
The casing seat can fracture on shut-in, causing gas to broach the sea floor and form a crater (worst case).

If an operator decides to divert the well to prevent possible crater formation or an underground blowout, the diverter system must be well designed to avoid failure. The operator must be prepared to give up all the possible benefits of shutting-in a well. Shutting-in the well requires the casing to be set at a sufficient depth so upward fluid migration will not occur and a crater will not form.

The shallow casing setting depth is a function of more than just the formation fraction pressure. Whether an underground blowout will stay underground cannot be predicted by fracture pressure alone. Planning the casing setting depth must also take into consideration the possible upward fluid migration and crater mechanisms of each particular area.

FLUID MIGRATION

Closing the well or restricting the fluid flow in the choke lines will increase the pressure in the well. If the pressure in the well reaches a limiting value, the cement bond or the formation around the well may fail. In any type of failure, a flow path may form for the high-pressure fluid. 1

CEMENT BOND FAILURE

Upward fluid migration through cement channels has caused numerous blowouts. Initial, small amounts of gas seeping around the casing can erode the cemented annulus between the borehole and casing. Thus, the cement job is one of the most important operations in drilling.

One of the main functions of cement placed in the borehole/casing or casing/casing annulus is to provide hydraulic isolation to prevent possible fluid migration. Cement channeling must be minimized or eliminated for effective isolation. The petroleum industry has therefore exerted a great deal of effort to solve channeling problems; however, the channeling mechanisms are poorly understood. Although a variety of solutions to the problem have been proposed, none has been consistently successful.2

ROCK FAILURE

Rock failure caused by the flow of high-pressure formation fluid can be basically divided into two categories: failure from tensile stress (tensile load), and failure from shear stress (compressive loads).

HYDRAULIC FRACTURE

Rock strength is a function of its structure, compaction, and type. Rock tensile strength varies in both the vertical and horizontal directions. The forces holding the rock together are the strength of the rock itself and the in situ stresses on the rock.

High-pressure fluid, resulting from the well control operation, in a well bore generates an hydraulic force on the well bore wall and in the pore spaces of the rock. If the pressure increases, the force applied by the fluid pressure on the rock will become equal to the forces holding the rock together. Any additional pressure will cause the rock to split or fracture. 3

Thus, from a macroscopic perspective, hydraulic fracturing occurs when the minimum effective stress at the well bore becomes tensile and equal to the tensile strength of the rock. 4 The fracture extends if sufficient pressure is applied by the injection of additional fluids. 3 5 Fracture propagation is a function of several factors:

In situ stresses in different layers of rock
Relative bed thickness of formations in the vicinity of the fracture
Bonding between formations
Mechanical rock properties (elastic modulus and Poisson's ratio)
Fluid pressure gradients in the fracture
Pore pressures in different zones. 6

Local stress fields and variations in stresses between adjacent formations are often considered the most important factors dominating fracture orientation and fracture growth. Production logs and other evaluation techniques have indicated that hydraulic fractures often terminate before propagating far into the bonding, impermeable layers (usually shale).

Soft, clay-rich materials normally have high stresses and often act as bonding layers. 7 8 Except for very soft formations, most formations have experienced some type of hydraulic fracturing. Fracturing operations have been successful in sand, limestone, dolomitic limestone, dolomite, conglomerates, granite washes, hard or brittle shale, anhydrite, chert, and various silicates. The plastic nature of certain soft shales and clays, however, makes them difficult to fracture. 3

An hydraulic fracture will usually propagate perpendicular to the direction of the minimum principal stress. 6 8 9 The direction of this stress will therefore determine if a fracture will be vertical or horizontal.

For well control considerations, hydraulic fracturing may lead to the serious risk of an open channel for upward fluid migration. If local stress conditions are likely to produce a horizontal fracture, an underground blowout is likely to stay underground.

If local stress conditions are likely to produce a vertical fracture and the permeability of the rock matrix surrounding the fracture is not great enough to dissipate the high pressure, the pressured fluid may migrate up through the fracture.

SHEAR FAILURE

Shear stress can cause rock failure if an impermeable formation overlays a permeable formation. In this case, the flow of high-pressure formation fluid can cause massive shear failure in the permeable formation before the overlying impermeable stratum is fractured. The consequences of such massive failure include increased sand production from the shear-damaged permeable formation, increased penetration rate when these strata are drilled, and compaction of these intervals. 1

FAULT PLANES

Existing fault planes crossing impermeable and sealing layers have allowed upward fluid migration that ended in the formation of craters. 1 10 11 Flow through the fault planes depends on such factors as the normal stress in the fault planes and the permeability of the sediments filling the fault plane.

The high-pressure fluid can wedge open a fault plane at a pressure below that which will cause fracture of the sealing layer. The permeability increases because of induced shear diligence within the fault plane by the high pressure. 1

CRATER MECHANISMS

Gas or liquid flowing at high velocity around the surface casing can erode shallow formation layers, forming a crater.

A significant erosion of the borehole wall cannot only create a crater but also lead to an excessive pressure drop in the well. This pressure drop can allow an additional flow of formation fluids (normally water) into the well. Erosion of the shallow formation by fluid flow has not been addressed by blowout-related literature, but the topic has been studied in various civil engineering problems, such as the erosion of river bottoms by moving fluids (water and solids).

Experiments have shown that erosion caused by fluid flow is a function of the fluid velocity and shear stress (from friction) at the eroding surface. 12 13 The greater the velocity and shear stress, the greater the erosion.

The erosion rate (mass of eroded material divided by time) is minimum and constant up to a certain value of velocity (critical velocity) or shear stress (critical shear stress). 12 13 For velocities greater than the critical value, however, the erosion rate increases rapidly as velocity increases. Erosion simulations have shown that as the eroded well bore diameter increases, the fluid velocity drops, and the erosion rate drops,

Thus, craters formed because of borehole erosion are normally small. The final diameter is a function of the formation erosion resistance and the fluid type. Figs. 1-3 are simplified views of a crater formed by borehole erosion.

In Fig. 1, the well was shut in after a gas kick. A leak in the casing allowed gas to migrate toward the surface casing shoe. The cement bond failed, allowing gas to erode the shallow formations. Figs. 2 and 3 are simulations showing the increased well bore erosion over time.

FORMATION LIQUEFACTION

Liquefaction is basically the process of making or becoming liquid. Formation liquefaction (also known as quicksand or boiling) occurs when the vertical effective stresses vanish; hence, the shear strength of cohesionless soils in the liquefied state is zero. 14-21 Therefore, zero vertical effective stress, defined as the overburden pressure minus a fraction of the pore pressure, may be used as the formation liquefaction criterion for cohesionless material.

Formation liquefaction can occur because of seepage where the overburden stress is small and the hydraulic gradient is high. 14 17 22 When formation fluid (gas, water, or both) flows through the soil, the fluid loses pressure (and energy). A drag effect occurs on the soil particles. If the drag effect is in the same direction as gravity, then the effective stress is increased, and the soil is stable. The soil may even become more dense.

Conversely, if water flows toward the surface, the drag effect acts against gravity thereby reducing the effective pressure between particles. If the velocity of the upward flow is great enough, particles can be buoyed up, reducing the effective stress to zero. The weight of the submerged soil is balanced by the upward-acting seepage force. This critical condition is commonly referred to as a sand boil or quicksand (liquefaction of sand deposits). The water pressure gradient at this point is the critical pressure gradient."

Quick conditions because of seepage forces are often found in excavations made in underwater fine sands subjected to upward fluid flow. As the velocity of the upward seepage force increases beyond a certain critical gradient, the soil boils more and more. At this point, structures fail by sinking into the quicksands

Quick conditions can also develop in a layered soil sequence composed of individual beds with different permeabilities. In this case, hydraulic conditions are particularly unfavorable where water initially flows through a very permeable layer with little loss of pressure. A quick condition can develop in shallow sand because of the increase in pore pressure. 14

PIPING

During an underground blowout, if the formation fluid reaches a cohesive soil layer, a phenomenon called "piping" or "tunnel erosion" may occur.

As the formation fluid flows through the soil, energy is transferred to the soil. Piping is the soil erosion as the seepage force acts on the soil. When formation fluid with sufficient velocity percolates through heterogeneous soil masses, the fluid moves preferentially through the most permeable zones and emerges from the ground as springs.

Piping refers to the erosive action of such springs where sediments are removed by seepage forces, thus forming subsurface cavities and tunnels.

For piping to occur, the soil must have some cohesion. The greater the cohesion, the wider the tunnel. 14 For piping to occur in cohesive materials such as clay, a crack or flow channel must be present to allow a concentrated fluid flow to develop. This crack could occur because of fracturing. 23 In the piping process, the formation fluid must be moving in the soil with sufficient volume and velocity to transport clay particles. This flow may be in a supersaturated layer with an under layer of impermeable material or along cracks in relatively, impermeable Soil. 24

Piping may also develop by backward erosion; the soil erosion progresses toward the fluid source. 14 If soil erosion because of piping reaches a critical value, entire structures can collapse from lack of support.

CAVING

Caving is the collapse of solids in and around the well. The borehole wall can collapse from shear failure because of a reduction in the hydrostatic pressure in the well bore or from tensile failure because of excessive production rates.

Caving from shear failure can be understood by analyzing the origin of the stress concentration at the well bore wall. Underground formations are exposed to vertical and horizontal compressive stresses that generally are not fully compensated by the drilling fluid pressure after the well is drilled.

For elastic formations, the load originally carried by the removed rock is partially transferred to the formations surrounding the borehole, creating a stress concentration around the borehole. The stress concentration generally does not present a problem if the well is drilled through competent rock. The stress concentration in weak rocks or in some shale sections, however, can lead to borehole failure.

Sand production often indicates caving from a high production rate. This type of caving is related to fluid drag forces on the formation grain.'

Although these two mechanisms probably act simultaneously, shear failure may be responsible for the production and removal of large amounts of sand and silt. Sand production from tensile failure in general tends to be less of a problem; as the cavity around the well bore grows, the fluid pressure gradient becomes smaller, leading to a drop (or complete stop) in sand production.

Migration of fines, such as clay, can block the pore spaces. The permeability of the surrounding formation decreases, increasing the drag forces. The higher drag forces may then initiate sand production. 4

Sand and silt production further complicate blowouts because of the increased wear and erosion of valves, diverter lines, and blowout presenters. Additionally, the excavation (if a permeable layer can help collapse the overlying sediments. 4

One documented case of a cratered well mentions that the material expelled from the crater formed a deposit approximately 40-in. thick at the edge of the crater and covered an area of about 99.7 acres. 25

HISTORICAL CASES

A number of cratered wells occurred in the early 1900s in Arkansas and Louisiana gas fields.

A typical cratered well was drilled in a region with shallow unconsolidated formations and shallow artisan water sand reservoirs. The water-filled sands were typically at depths of about 600-1,000 ft. The gas-producing reservoir was found at an average depth of 2,200 ft and averaged a 0.45 psi/ft pressure gradient.

One particular well was drilled, completed, and shut in. Soon thereafter, 550-psi was on the annulus between the 6-5/8-in. and 10-in. casings. Gas began to seep around the 20-in. surface casing. The well was opened to the atmosphere to release the pressure.

Water started to flow around the surface casing, and the formation caved. Attempts to kill the well by pumping mud and water into the casing were unsuccessful. A huge crater formed 5 days after the blowout. The derrick fell, and the well caught fire. The crater was 390-ft in diameter and approximately 100-ft deep. 25

A study of the incident revealed the most likely crater mechanisms were borehole erosion, formation liquefaction, and caving. These mechanisms probably occurred simultaneously and in the following manner:

Gas seeped around the well bore, indicating the welt bore was eroded and enlarged.
High pressure in the initially tight borehole/casing annulus caused gas to flow from the well bore into the shallow, poorly cemented formations. This flow through fractures or through the pore media was responsible for the liquefaction of the cohesionless soil and the poorly cemented rocks.
The enlargement of the well bore by gas erosion caused the pressure to drop in the well, with consequent production of water and sand.
The surrounding formations caved. High fluid turbulence at the surface also contributed to the enlargement of the crater.

DEEPWATER CRATER

In the deepwater Gulf of Mexico, several companies have encountered shallow (below the mud line) water flows.

For typical deepwater wells in the area, the water depth is 3,900 ft, and the first casing, the 30-in. pipe, is set and cemented at a depth of 4,140 ft, 240 ft below the mud line (BML). Several slightly pressured (8.7-9.3 ppg equivalent mud weight) and high permeable water zones were expected between 1,500 and 1,800 ft BML.

To isolate these water-flowing zones, the 20-in. casing is normally set and cemented at 2,006 ft BML.

The fracture pressure gradient, obtained from nearby wells, is close to 8.8 ppg at the 30-in. casing shoe. Formations at these shallow depths are likely to be very heterogeneous. Any flow from slightly high-pressure, deep sands is likely to go around the 30-in. casing shoe.

On one such well, water initially began to flow through the return ports to the 30-in. x 20-in. annulus. The exhaust ports plugged, causing the flow to go around the 30-in. shoe and to the sea floor. A small 40-ft circular crater formed around the wellhead (Fig. 4). The most likely crater mechanism in this case was piping.

REFERENCES

Walters, J.V., "Internal Blowouts, cratering, Casing Setting Depths, and the Location of Subsurface Safety Valves," SPE Drilling Engineering, December 1991, pp. 285-92.
Lockyear, C.F., Fyan, D.F., and Gunningham, "Cement Channeling: How to Predict and Prevent," SPE paper 19865, prepared for the SPE Annual Technical Conference and Exhibition in San Antonio, Oct. 8-11, 1989.
Prior, D.B., Doyle, E.H., and Kaluza, M.J., "Evidence for Sediment Eruption Deep Sea Floor, Gulf of Mexico," Science, Vol. 243, Jan. 27, 1989, pp. 517-19.
Fjaer, E., Holt, R.M., Horsud, P., Raaen, A.M., and Risnes, R., Petroleum Related Rock Mechanics, Elsevier, New York, 1992.
Haimson, B., and Fairhurst, C., "Initiation and Extension of Hydraulic Fractures in Rocks," SPE paper 1710, 1967.
Veatch, R.W., Moschovidis, Z.A., and Fast, C.R., "An Overview of Hydraulic Fracturing," SPE monograph Vol. 12, 1989, pp. 1-3.
Harrison, E., Kieschnick, W.F., and McGuirre, W.J., "The Mechanics of Fracture Induction and Extension," Transactions of the AIME, Vol. 201, 1954, pp. 252-63.
Warpinski, N.R., and Teufel, L.W., "Influence of Geologic Discontinuities on Hydraulic Fracture Propagation," SPE paper 13224, 1984.
Warpinski, N.R., and Smith, M.B., "Rock Mechanics and Fracture Geometry," SPE monograph Vol. 12, 1989, pp. 57-80.
Adams, N., and Thompson, J., "How a geothermal blowout was controlled," World Oil, June 1989, pp. 36-40,
Adams, N., and Kuhlman, L.G., "How to prevent or minimize shallow gas blowouts-Part 1," World Oil, May 1991, pp. 51-58.
Gaylord, E.W., "A Laboratory Study of the Effects of Hydraulics on Hole Erosion," SPE paper 12118, 1989.
Kimphuis, J.W., and Hall, K.R., "Cohesive Material Erosion by Unidirectional Current," Journal of Hydraulic Engineering, ASCE, Vol. 109, No. 1, 1983, pp. 49-61.
Bell, F.G., Engineering Properties of Soils and Rocks, Butterworths, London, 1983.
Clough, G.W., Iwabuchi, J., Rad, N.S., and Kuppusamy, T., "Influence of Cementation on Liquefaction of Sands," Journal of Geotechnical Engineering, Vol. 115, No. 8, 1989, pp. 1102-17.
Committee on Soil Dynamics of the Geotechnical Division, "Definition of Terms Related to Liquefaction," Journal of the Geotechnical Engineering Division, American Society of Civil Engineers, 1978, pp. 1197-1201.
Lee, I.K., White, W., and Ingles, O.G., Geotechnical Engineering, Pitman Books Ltd., London, 1983.
Scott, C.R., An Introduction to Soil Mechanics and Foundations, Maclaren & Sons, England, 1963.
Seed, B.H., Idriss, and Arango, I., "Evaluation of Liquefaction Potential Using Field Performance Data," ASCE Convention and Exposition in St. Louis, Oct. 26-31, 1981.
Terzaghi, K., and Peck, R.B., "Soil Mechanics in Engineering Practice," John Wiley & Sons Inc., 1987.
Zeevaert, L., Foundation Engineering for Difficult Subsoil Conditions, Van Nostradamus Reinhold Co., 1973, pp. 175-215.
Kramer, S.L., and Seed, H.B., "Initiation of Soil Liquefaction Under Static Loading Conditions," Journal of Geotechnical Engineering, Vol. 114, No. 4, 1988, pp. 412-30.
Ghuman, O.S., Allen, R.L., and McNeill, R.L., "Erosion, Corrective Maintenance, and Dispersive Clays," Dispersive Clays, Related Piping, and Erosion in Geotechnical Projects, ASTM special technical publication 623, 1977, pp. 172-90.
Crouch, R.J., "Tunnel-Gully Erosion and Urban Development: A Case Study," Dispersive Clays, Related Piping, and Erosion in Geotechnical Projects, ASTM special technical publication 623, 1977, pp. 58-73.
Hills, H.B., "Crater Wells Richland Gas Field Louisiana," Technical Paper 535, U.S. Bureau of Mines, 1932.
Martinez, S.J., Steanson, R.E., and Coulter, A.W., "Formation Fracturing," SPE reprint series No. 28, 1990.