KILL OPERATION REQUIRES THOROUGH ANALYSIS

L. William Abel
Wild Well Control Inc.
Spring, Tex.

Full control of a blowout well requires a properly designed post-capping kill operation because failures in regaining well control usually occur during the kill operation, not during capping.

Capping (the installation of pressure control or diverter equipment on the wellhead) is generally very reliable in gaining control of a blowout well.

The following techniques are some of the viable means of killing blowout wells once the capping assemblies are in place:

Direct shut in of the flow
Bullheading
Momentum kill
Volumetric control for migration of fluids or lubrication after migration ceases
Dynamic kills (friction-based dynamic kills or mass flow rate kills).

The objective of most post-capping operations is to stop the flow and put the well under hydrostatic control (Table 1(58261 bytes). The means of killing a blowout once capping assemblies are in place should be chosen with care to avoid problems such as cratering, equipment failure, and underground blowouts. The particular circumstances and well integrity will dictate which kill method will be the most viable.

DIRECT SHUT IN

Shut in of flow is defined here as the technique of capping a well and stopping the flow by either directly closing the blowout preventer (hard shut in) or choking back a diverted flow (soft shut in).

Many blowout wells have been capped and then directly shut in for successful control. Of all the kill methods available, this method usually results in the highest shut-in pressure and consequently can be the most stressful for the wellhead, downhole equipment, and formation.

This high shut-in pressure may or may not be of concern depending on the integrity of the well equipment and exposed formations.

The well can be shut in by either directly closing a blind ram (a hard shut in as on the majority of the wells in the Kuwait well control project) or by diverting the flow through a choke and then closing in the flow gradually (Fig. 1 (107303 bytes).

A pressure profile (or load line) in the well is produced as the pressure builds up following shut in. The pressure depends mainly on the strength and characteristics of the reservoir, the type of fluids in the well bore, and most importantly, the method used to shut in the well. Assuming that no pumping occurs, the maximum stress can be expected if the well is allowed to build up to its maximum shut-in surface pressure. Fig. 2 (46416 bytes) shows pressure profiles for flowing conditions, shut in conditions after the reservoir has reached its static pressure, the burst load line for casing, and fracture of the formation.

The well in Fig. 2 (46416 bytes) has been designed so the casing and formation will withstand shut in against a full column of gas to the surface; therefore, the well can be closed in safely after capping. The dynamics of the shut-in procedure must be examined with consideration for the integrity of the well bore, downhole tubulars, and surface control equipment.

In a blowout well there can be damage from flow cutting, temperature effects, or dropped pipe.

Additionally, water hammer effects may need to be examined. If the well is producing at a low gas/oil ratio (GOR), water hammer can cause pressure transients that can damage the casing and fracture the formation. Closing-in against the choke may be considered if there is any doubt of the well integrity, particularly for high pressure wells and for wells with low GOR flow streams. The water hammer effect is also affected by damping, which depends on the following:

Compressibility and density of the produced fluid
Velocity of the fluid at the time of closure
Time for closure of the surface control device
Elasticity of the formation and casing
Geometry of the hole.

Fig. 3 (48915 bytes) shows the water hammer pressure transients traveling down the well bore from the point of origination at the blowout presenter (BOP). The amplitude or intensity diminishes from damping.

One might expect that a short-duration pressure transient would be acceptable and present no danger of causing damage to the casing integrity, depending on the failure mechanism of the casing (ductile or brittle). This judgment should be made on a case-by-case basis, however.

In an open hole section, pressure transients need to be analyzed in consideration of the failure mechanics of the rock. If loss of circulation can only occur by fracturing the rock and not from a matrix injection, the transient will cause a minor fracture. After the pressure transient passes or dissipates, the fracture will close, and the original rock integrity will be regained.

If the pressure transient causes a breakdown of filter cake, which cannot heal quickly, loss of returns can be expected. In some cases, a loss of returns can result in an underground blowout.

If shut in (either hard or soft) causes the pressure profile to exceed the formation integrity, an underground blowout may occur, and other kill methods should be considered. Fig. 4 (38871 bytes) shows a situation in which either a hard or soft shut in would be unacceptable because the shut-in pressure profile will exceed the formation integrity.

BULLHEADING

In some cases, bullheading (pumping fluid into the well from surface) will be the most expedient manner to bring the well under control. Bullheading may be necessary because it may not be possible to circulate conventionally because of the downhole configuration or damage. When properly done, bullheading is efficient.

Care should be taken beforehand to ensure that no further damage is caused by the stresses bullheading induces. Generally, bullheading will exert more pressure on the well than circulation from the bottom or a dynamic kill.

Prior to starting the bullhead procedure, the well will either be on diversion or shut in. If the well has been shut in for some time, the wellhead pressure will usually be at its highest value. Additionally, the initial pumping pressure will have to be greater than the shut-in pressure to force the fluid into the well. This higher pressure will add stress to the downhole equipment, open hole interval, and surface equipment.

Providing the tubulars in the well and surface equipment can withstand full shut-in pressure and the additional pressure exerted by bull-heading, the well can be bullheaded dead by pumping in kill weight mud at the appropriate rate.

The most important factor is whether fluid can be forced into the formation. In some cases, bullheading can risk loss of circulation, thus complicating the control operation. Fig. 5 (45160 bytes) compares pressure profiles for shut in, buildup, and then a pump-in operation (Curve A) and bullheading just after shut in (Curve B).

In most cases where bullheading is considered, less surface pressure will result if the procedure is begun immediately after shut in and before the well builds up pressure. The pressure produced during bullheading depends on several factors:

Well geometry (OD, ID, deviation, etc,)
Rate of buildup (flow rate)
Properties of the flow (compressibility, Z factor, temperature, density, etc.)
Pump rate for the mud introduced into the well
Density and rheology of the mud pumped,

These factors are complex and do not readily yield to generalizations and rules of thumb; thus, a computer-aided technique is recommended to guide the bullhead operation.

If pumping in immediately after well shut in is the chosen method, one must determine an optimum pump rate. In most blowouts, the downhole conditions may not be known, or it may be impossible to determine casing or wellhead integrity. Questions of damage from flow or heat usually cannot be addressed until the well is killed. The blowout intervention team then has difficult decisions in trying to avoid further complications and damage.

In some rare cases, catastrophic effects have occurred in post-capping kill operations. In one case, an entire platform collapsed after a bullhead attempt. Extreme caution and careful study are warranted in killing a well that has blown out and is on diversion. Many wells have been successfully killed by bullheading, however.

One goal of the bullhead operation should be to minimize the surface pumping pressure. Downhole stresses will then be minimized.

If the well is on diversion, one must study the procedures, kill rate, mud density, and Theological properties and compare these to the well parameters. Computer-aided techniques are recommended.

The analysis should include the following parameters:

Hole geometry
Initial flowing wellhead pressure and any buildup before the job commences
Reservoir parameters (compressibility, Z factor, temperature, buildup rate, reservoir pressure, etc.)
Estimate of leakoff to the formation
Pill to be pumped ahead (volume, weight, and viscosity)
Kill mud weight and viscosity
Tube efficiency for friction calculation
Minimum and maximum kill rate considered.

The calculations take into consideration the compressibility of the well fluid (gas), gain in density of the kill fluid, and the frictional losses of the fluids pumped:

WHP = (p2 + Br) + APmud + APpill - HPmud - HPpill

In this equation, WHP is the wellhead pressure; P2 is the pressure in the gas column from compression of the column volume; Br is the pressure buildup in the gas column from the well flowing into the well bore; DP is the frictional pressure losses from fluid movement; and HP is the hydrostatic pressure from the length of fluid column pumped in (Fig. 6 (30806 bytes).

MINIMUM KILL TECHNIQUE

If a bullhead procedure is to be used, there is an optimum rate that will produce the minimum stress in the well. Pumping at a very slow rate achieves nearly the same effect as allowing the well to build up to maximum pressure and then bullheading. This procedure will increase wellhead pressure.

Thus, a high pump rate will have a better chance of minimizing the wellhead pressure. At very high pump rates, frictional pressures will have a notable effect and will be reflected to the surface. Therefore, there is an optimum rate that will produce the lowest surface pumping pressures, considering these factors. Killing the well at optimum rate is referred to as the minimum kill technique.

Fig. 7 (55102 bytes) shows the graphical output from a spreadsheet program modeling a bullhead kill. In this case, it should be possible to keep the surface pressure less than 500 psi if the pump rate is 95 bbl/min. This rate may seem excessive, but there will only be surface lines to pump through and very little pressure to overcome, so the horsepower requirements will not be excessive. If a limit of 1,000 psi wellhead pressure were chosen, the pump rate could be reduced to 65 bbl/min.

MOMENTUM KILL

The momentum kill technique uses the velocity and density of a pumped kill fluid to stop a blowout flow (Fig. 8 (50861 bytes). It is essentially a bullhead technique but the unique aspect is that a wellhead or pressure control equipment are not necessary.

This technique has been used in shallow wells (

Density and pump rates are determined by setting the momentum of the blowout fluids (Qr) equal to the momentum of the kill fluid (Qk) and solving for either kill pump rate or density. One must be cautioned that this approach only yields the rate to cause the reservoir fluid to be stopped at the point where the kill and reservoir fluids meet. What happens thereafter is essentially a bullhead technique, where a "liquid valve" is established at the end of the stinger (or where the two fluids meet).

The momentum of the fluid is dependent on the flow area. The momentum of the kill fluid inside the stinger (Point I on Fig. 8(50861 bytes)) is much higher than that in the borehole (Point 2 on Fig. 8(50861 bytes)), which can lead to underpredicting the pump rates necessary. 5-7 Point 2 is the proper point to consider for the balance point for calculation of momentum forces. This approach will require higher pump rates or density but has been proven to be valid in actual kills in the field.

VOLUMETRIC WELL CONTROL

Well control can be accomplished without circulation with the volumetric method. The volumetric method involves bleeding mud from the system to allow for gas expansion and migration while maintaining control of the well. The method is essentially a constant bottom hole pressure technique.

The objective is to prevent additional influx into the well bore without exceeding the fracture pressure of any exposed formations or the equipment pressure limits. The method is particularly useful if the string is plugged or if the well cannot be circulated.

If bullheading is done and at the end of the job there is gas left in the well bore, volumetric methods may be required to complete the kill operation. One such case is where the string is dropped in the hole and bullheading kills the annulus but not the inside of the string. Any gas trapped in the string will migrate up the hole and will need to be controlled volumetrically.

If the well is closed in with both kill and reservoir fluid in the column, volumetric methods can be used to allow the reservoir fluids to migrate up the hole and to be replaced with kill mud. This procedure can be done without a kill string for circulation. Fig. 9 (65481 bytes) shows the general sequences for controlling the well while handling gas migration.

The test well at Louisiana State University was used to demonstrate this method and to train personnel in its application. In an exercise, natural gas was injected at high pressure into the bottom of a 7,500-ft well with tubing (drill pipe) to surface.

The participants in this exercise were allowed to see the annular pressure only and had the option of observing surface pressure or bleed-back mud. A surface reading of the shut-in drill pipe pressure (Sidpp) is available to the instructor and is used to determine bottom hole pressure (BHP). The BHP is the hydrostatic head of fluid column in the drill pipe plus the Sidpp.

In using the shut in casing pressure (SICP) to predict BHP, one must use a systematic procedure to calculate the pressures in the well bore.

Fig. 10 (46376 bytes) shows the results of a volumetric control exercise, with BHP calculated from annular pressure and the worksheet approach and from the observed Sidpp. Both methods yield identical results. The pressure exerted downhole varied from about 100 psi to a high of 400 psi.

From Fig. 10 (46376 bytes), one can see that the BHP is controlled by the bleed off of mud and can be altered or maintained. Time data Points 2-5 show migration without bleeding off mud. Time data Points 6-10 show that a constant BHP can be maintained. The exercise demonstrated that the BHP can be controlled to a value above a set amount (reservoir pressure in this case was 2,900 psi) and below a maximum (in this case 3,400 psi) without circulation.

LUBRICATION

After all the formation fluids have migrated to the surface, pressure increases will cease. The well can be bullheaded, but this may force the formation fluids down the hole and not back into the formation. Migration will then recur, complicating control efforts. The formation fluids can be lubricated out and kill fluid placed in the well.

The lubrication method is the second phase of the volumetric control technique. Kill fluid is pumped into the well in small increments (size depends upon pressure limitations, open hole integrity, etc.). The kill fluid and formation fluid are given time to change places, and the formation fluid is bled off in small increments. ,

The increment to bleed is guided by the pressure response seen during the bleeding process. This method is essentially a constant bottom hole pressure method. Properly done, all the formation fluid will be replaced with kill-density fluid. The well will be filled with kill fluid, and thus be statically under control.

DYNAMIC KILL

Dynamic kills have been accomplished with relief wells and with workstrings in the blowing wells. These operations fall into two broad categories: friction and mass flow rate kills.

FRICTION KILL

The friction kill uses a large pad of water ahead in the pump schedule. The pumping rate is brought up to the value that will control the bottom hole pressure with the combination of water hydrostatic pressure, friction created in the flow path, and choking pressure at the exit of the flow (Fig. 11(50797 bytes). The friction method of dynamic kill works well and does not require complex calculations. The drawbacks are that massive hydraulic horsepower may be required to obtain the rate necessary to effect the required frictional losses in the blowout well. Additionally, a conduit to the productive horizon is necessary. This conduit can be a snubbed in workstring, an existing workstring, or a relief well.

Snubbing on a blowout well can be time consuming and expensive, with no assurance of success. Relief wells are also expensive and time-consuming projects. The pure friction method uses basic assumptions and does not account for the reservoir characteristics; therefore, massive volumes of fluid may need to be pumped to overcome the blowout flow.

MASS FLOW PATE KILL

If the blowout well bore has a geometry that will allow frictional losses to be generated at reasonable pumping spreads and rates (10,000 hydraulic horsepower or less), a mass flow rate kill option should be considered (Fig. 12 (52362 bytes).

The mass flow rate kill option requires careful analysis of two-phase flow behavior, which requires computer assistance. Providing the blowout can be modeled accurately, the kill requirements in volume, hydraulic horsepower, and pumping spread can be significantly reduced.

The information required to perform the analysis can be estimated if precise information is not available. The mass flow rate approach can yield substantial savings in horsepower and pumping rates compared to the friction method.

REFERENCES

Craft, Holden, and Graves, Well Design, Prentice Hall Inc., New Jersey, 1962, P. 23.
Skeeter, V.L., and Wylie, E.B., Fluid Mechanics, McGraw Hill Books, New York, 1979, P. 239,
Wild Well Control Inc., Bullhead, Excel 4.0 spreadsheet program for predicting surface pumping pressure while bullheading a well, 1993.
Abet, L.W., Bowden, J.R. Sr., and Campbell, P.)., Firefighting and Blowout Control, Wild Well Control Inc., Spring, Tex., 1994, 504 pp.
Moore, P.L., Drilling Practices Manual, PennWell Publishing Co., Tulsa, 1986, pp. 544-45.
Grace, R.D., "Practical Considerations in Pressure Control Procedures in Field Drilling Operations," journal of Petroleum Technology, Society of Petroleum Engineers, August 1977.
Watson, W.D., and Moore, P.L., "Momentum kill procedure can quickly control blowouts," OGJ, Aug. 30,1993, pp. 74-77.
Blount, E.M., et al., "Dynamic Kill: Controlling Wild Wells a New Way," World Oil, October 1981.
Bourgoyne, A.T., and Wild Well Control Inc., DYN-X, Excel 4.0 spreadsheet program for predicting dynamic kill rates and two-phase flow behavior, 1993.