Well control procedures developed for multilateral wells

Nov. 4, 1996
Yuejin Luo BP Exploration Sunbury-on-Thames, U.K. Allan Gibson, Carl Mountford, Ted Hibbert BP Exploration Aberdeen Curtis Weddle BP Exploration Houston Method for calculating kick tolerance volumes in vertical, diviate, or horizontal wells [79473 bytes] For multilateral wells, conventional well control methods require some additional precautions and measures compared to well control methods used in single-bore wells. Multilateral well technology is increasingly used to improve well
Yuejin Luo
BP Exploration
Sunbury-on-Thames, U.K.

Allan Gibson, Carl Mountford, Ted Hibbert
BP Exploration
Aberdeen

Curtis Weddle
BP Exploration
Houston

For multilateral wells, conventional well control methods require some additional precautions and measures compared to well control methods used in single-bore wells.

Multilateral well technology is increasingly used to improve well productivity and reserves, as well as to reduce drilling and facility cost in field developments.

Kick tolerance is one of the important criteria in designing casing setting depth during well planning, and it is also an important parameter that must be monitored during drilling operations.

Several methods exist for calculating kick tolerance.1-4 A method that is used within BP Exploration is described in the accompanying box. The method calculates the maximum volume of formation influx that can be shut in and circulated out of the well without breaking down the formation at the open hole weak point.

This method is applicable to a single-bore well, though it may be vertical, deviated, or horizontal. In a multilateral well, if the hydraulic isolation between well bores is sufficiently competent, the method can be applied to any of the well bores being drilled. If this is not the case, however, it needs modifications:

  • The maximum allowable annular surface pressure (Maasp) should be calculated based on the weakest formation in all well bores.

  • Possible influxes from all well bores should be considered.

Based on the principles and the method described in the accompanying box, a new method can be derived for determining kick tolerance in multilateral wells. The procedure of the new method for a dual-lateral well is as follows:

1. Determine Maasp.

For an active well bore (containing drillstring), Maasp can be obtained by Equation 1.

For a static well bore (without drillstring), possible different mud weights in the active and static well bores should be considered. So, Maasp for a static well bore is determined by Equation 2.

The final Maasp that should be used for subsequent kick tolerance calculations is determined by Equation 3 (Maasp is the smaller of Maasp1 and Maasp2).

2. Determine the kick tolerance volume assuming an influx from the active well bore. The maximum allowable gas influx height in the active well bore is determined by Equations 4 and 5.

Based on Hmax-1, the kick tolerance volume VKT-1 for an influx from the active well bore can be calculated following Steps 3-6 in the box.

3. Determine the kick tolerance volume by assuming an influx from the static well bore. Possible different mud weights in the active and static well bores should be considered (Equations 6 and 7).

Based on Hmax-2, the kick tolerance volume VKT-2 for an influx from the static well bore can be calculated using the standard procedure.

To facilitate applications of the above procedure, a Microsoft Excel spreadsheet program was developed. Based on the well data in Fig. 1 [23917 bytes], an example calculation is shown in Fig. 2 [23917 bytes] and Table 1 [65458 bytes].

Kick prevention

Essentially all techniques used in conventional wells for preventing kicks can apply in multilateral wells. These concepts have been discussed in numerous publications. However, as lateral well bores are often drilled at high angles and in small hole sizes, some additional precautions should be taken:

  • The potential kick intensity can be high if there is a long, high-angle section through the reservoir formation.

  • The equivalent circulating density (ECD) is relatively high in a lateral well bore because of its smaller hole size and high deviation. There will be a greater bottom hole pressure reduction when circulation stops. Therefore, it is important to flow check the well when circulation stops (for making connections, etc.) to ensure that the well is stable without ECD effects.

  • When the string is pulled from a lateral well bore, the swab pressure will be relatively high (because of its smaller hole size). Therefore, it is important to ensure that mud rheology is conditioned prior to tripping out and that tripping speed is controlled.

Detection

All warning signs used in conventional wells for detecting a kick should be effective in multilateral wells. As in a conventional well, the two most important warning signs of a kick in a multilateral well are the pit gain and increased mud return flow. Pit gain is often more sensitive for a low-intensity kick (because of low underbalance pressure or low formation permeability). The increased return is often more sensitive for a high-intensity kick.

Special caution should be taken if there is a tendency of losses in any of the well bores, which may create a situation where mud is lost into a formation in one well bore while influx is taken from the other. This may obscure both pit gain and mud return. In this circumstance, the normal loss rate and pit level should be closely monitored and any change in the normal trends investigated.

Once a kick has been detected, a judgment should be made regarding whether the influx is taken from the active or the static well bore. In many cases, this judgment can be based on operation circumstances and knowledge of reservoir characteristics. But in some cases, making a judgment may not be so straightforward.

There are some warning signs that should be specific to an influx from the active well bore.

  • There is an increase in penetration rate (drilling break).

  • The shut-in casing pressure (SICP) is higher than the shut-in drill pipe pressure (Sidpp), when the active well bore is nonhorizontal. (This sign is valid only during the early stage when the influx is still below the lateral junction point.)

  • During the shut-in period when volumetric control is implemented to control gas migration, Sidpp is relatively stable while SICP increases. (This sign is valid only during the early stage when the influx is still below the lateral junction point.)

An influx from the static well bore can be confirmed by the following signs during the well shut-in period when the influx is still below the lateral junction point:

  • Sidpp is equal or close to SICP (when the active well bore is nonhorizontal).

  • Both SICP and Sidpp increase during gas migration until the gas reaches the junction point.

Based on the above warning signs, accompanied by the pit gain or increased mud return flow, it should be possible to make a judgment regarding which well bore the influx is taken from. To achieve this, it will be necessary to monitor the parameters during all phases of drilling operations.

Shut-in procedure

The shut in procedures have also been discussed in many publications. As multilateral wells are often high angle or horizontal, however, some additional points should be emphasized:

  • The fast shut-in method should be used upon detecting a kick to minimize the influx volume. Studies showed that the potential water-hammer effect associated with the fast shut-in is negligible.

  • The shut-in Sidpp will be close or equal to the SICP if a kick is taken in a high angle of horizontal section. This is because the kick only causes a small or no reduction in the annular hydrostatic pressure.

  • Zero shut-in pressures (Sidpp and SICP) do not mean there is no kick. Together with a positive pit gain, it may indicate that the kick is still in the horizontal hole section that may be caused by swabbing or improper hole fill up on trips.

  • The conventional method that determines the influx density/type (gas/water/oil) based on pit gain, Sidpp, and SICP cannot be applied if the kick is taken from a high angle or horizontal hole section. There is no simple alternative method for field applications. However, a gas influx can be recognized by the continuous increase in the casing pressure due to gas expansion above the horizontal hole section. This may be caused by gas migration during shut in or by mud circulation.

  • During the well shut-in period, free gas usually migrates up the annulus if the hole angle is less than 90°. The migration rate will depend upon mud rheology, hole angle, and hole size. Increasing mud yield stress or gel will reduce the gas migration rate.

  • The migration rate should not be calculated based on the increase in SICP, as it may underpredict the migration rate.

  • Gas does not migrate if the hole angle is 90° or higher, if the gas is dissolved in oil-based mud, or if the gas is trapped as small bubbles in mud by its gel strength/yield point.

Kill procedure

Once the well has been shut in upon detecting a kick, a decision must be made regarding the most appropriate action to kill the well.

As in a conventional single well-bore well, an attempt should always be made to use one of the standard kill techniques (wait and weight or driller's method). This is particularly true when the influx is taken from the active well bore, and in this case the standard procedures and kill sheet can be used in the same way as in a single-bore well. Due consideration should be taken for the additional stress imposed on the static well bore during the shut-in and kill operations.

In cases where the influx is taken from the static well bore, or there is considerable uncertainty regarding which well bore it is taken from, the procedures and the kill sheet for implementing the standard techniques need to be modified.

In the standard techniques, the kill mud weight is calculated based on Sidpp and true vertical depth (TVD) of the hole (Equation 8). Using this technique, the well is killed when the kill mud has been circulated into the hole and returned to surface.

In a multilateral well, however, if the influx is taken from the static well bore that contains no drillstring, the annulus from the junction point to the kick zone can not be circulated to the kill mud. Therefore, if the above technique is used, there will be a remaining shut-in surface pressure after the kill, which can be calculated by Equation 9.

In this case, other unconventional techniques (such as bull-heading, etc.) may have to be used, or the drillstring has to be stripped or snubbed into the static well bore (if possible) before the well can be killed. This will complicate the kill operation considerably.

To simplify the kill operation, an alternative is to circulate the hole to a higher kill weight than that used in the standard technique. One option is to calculate the kill mud weight based on the depth at the lateral junction point (Equation 10).

If this kill mud is used, the surface pressure should return to zero when the kill mud returns to surface, which would allow the drillstring to be tripped into the static well bore. To implement the technique, the kill procedure will be the following:

  1. Calculate the first kill mud weight using Equation 10, and prepare the corresponding pump pressure schedule using an appropriate kill sheet.

  2. Circulate the well to the first kill mud. When it returns to surface, the surface Sidpp and SICP should be zero. If not, repeat Steps 1 and 2.

  3. Pull the string from the active well bore, and trip into the static well bore (with great caution). As the string is tripped into the influx, it will be displaced above the bit which may induce surface pressure again. So, monitor the well closely and be prepared to shut in and circulate. Once the influx above the bit has been circulated out, the blowout preventer can be opened to continue tripping in. This may have to be repeated before the bit reaches hole bottom.

  4. Once the string is on bottom in the static well bore where the influx is taken, the well should be shut in and circulated to the second kill mud weight calculated by Equation 11.

  5. Trip back into the original active well bore, and circulate to the above mud weight plus a safety margin.

The kill operation is then complete.

It should be pointed out that by using the above procedure, the well bores will endure higher-than-necessary pressures during the first kill. So a check must be made to ensure that the formation will not break down at any weak point. If it is likely, then the first kill mud weight should be reduced to the maximum safe value. In that case, there will be remaining shut-in surface pressures when the first kill mud has returned to surface from the active well bore, and special techniques may have to be used for the final kill.

These procedures have been summarized in a flow chart shown in Fig. 3 [45972 bytes].

To facilitate the implementation of this kill technique, an Excel spreadsheet program was developed. An example calculation is shown in Fig. 4 [55798 bytes], based on the well data in Fig. 1.

Weak casing window joint

Currently, some multilateral systems use a casing joint with a premachined window for kicking off the lateral well bore. The joint has a lower burst and collapse strength than the main body of the casing string. This joint can be a weak point if an influx is taken when casing is run.

The volume of influx that can be taken without collapsing the casing window joint can be calculated based on its collapse strength. The worst-case scenario occurs when the casing window joint is just below the blowout preventer when the influx top arrives at the wellhead. This results in the maximum influx volume in the well bore and imposes the maximum pressure on the casing window joint.

Based on its collapse strength, the maximum allowable influx height during running the window joint to any point of the well can be calculated by Equation 12. The corresponding influx volume can be calculated by Equation 13.

Based on Boyle's law, the above volume can be converted to the shut-in volume (when the influx is at bottom) by Equation 14.

An example calculation using an Excel spreadsheet program is shown in Fig. 5 [23716 bytes].

Acknowledgment

The authors wish to thank BP Exploration for permission to publish this article and Dr. Paul Lurie for his numerous comments during the course of this work.

References

1. Pilkington, P.E., and Niehaus, H.A., "Exploring the myths about kick tolerance," World Oil, June 1975.

2. Redmann, K.P. Jr., "Understanding kick tolerance and its significance in drilling planning and execution," SPE Drilling Engineering, December 1991.

3. Wessel, M., and Tarr, B.A., "Underground flow well control: The key to drilling low-kick tolerance wells safely and economically," SPE Drilling Engineering, December 1991.

4. Parfitt, S.H.L., and Thorogood, J.L., "Application of QRA methods to casing seat selection," Paper 28909, presented at the Society of Petroleum Engineers European Petroleum Conference, London, Oct. 25-27, 1994.

The Authors

Yuejin Luo works in shared petrotechnical resource, BP Exploration, in Sunbury. He is currently working in the areas of well control, hole cleaning, and hydraulics. Previously, he worked at the Petroleum Institute of Exploration & Production in Beijing as a drilling engineer for 4 years.

Luo earned a BS in petroleum engineering from the University of Petroleum in China in 1981, and a PhD in petroleum engineering from Heriot-Watt University in the U.K. in 1988.

Allan Gibson is a drilling engineer with Dresser Drilling & Production Services. Gibson is currently engaged in deepwater drilling on BP Exploration's Foinaven field, the first oil development in West Shetland. Previously, he held drilling engineering and supervisory positions both on and offshore for Unocal, Agip, Premier Consolidated, and Charthouse Petroleum. He holds a BS in mechanical engineering from Glasgow University.
Ted Hibbert is a senior drilling engineer with BP Exploration. He is responsible for well construction activities in the Forties field. Previously, he held drilling engineering positions onshore and offshore in both the Middle East and the North Sea. He has also held drilling engineering positions in subsea activities on the Don, Magnus, Buchan, and Clair fields. He holds a BS in mechanical engineering from Imperial College in London.
Curtis E. Weddle III is a global drilling consultant responsible for well control at BP Exploration, where he has worked for the last 14 years. He earned a BS degree in civil engineering from Oklahoma State University and is a registered professional engineer in Oklahoma. In a number of assignments for major and independent oil companies, Weddle has drilled in most of the U.S. hydrocarbon bases. Recently, he has worked as a commercial analyst for BP and in his current assignment has worked on BP projects worldwide.
Carl Mountford is the senior drilling engineer for BP Exploration's Schiehallion development, West of Shetland. After working in deep, high-pressure, semisubmersible drilling operations in the North Sea, he supported BP's U.K. continental shelf and overseas developing assets based in London.
From 1990 to 1993, he was seconded to work in drilling operations for Abu Dhabi Marine Operating Co. (ADMA OPCO) in the U.A.E. Since returning to the U.K., he has worked in development drilling engineering support of U.K. continental shelf assets and specifically the new West of Shetland fields.

Copyright 1996 Oil & Gas Journal. All Rights Reserved.