Multilateral technology gains broader acceptance

MULTILATERAL WELL DESIGN-1

Mike R. Chambers
Mobil E&P Technical Center
Dallas

The implementation of multiple well bores, or multiple horizontal wells exiting a single well bore, has gained wider acceptance in the oil industry, particularly from a reservoir-management point of view.

This first of a two-part series describes the history, terminology, and reasons for using multilateral technologies. The second part will describe the classification system and completion equipment selection criteria.

Multilateral history

The application of drilling multiple well bores or laterals from a single well is not a new concept with evidence indicating such endeavors began in the 1920s. In 1928, E.A. Spencer filed a patent for a tool designed to cut multiple windows in casing with the use of a whipstock-type device. In 1931, D. Dana also filed a patent for a downhole splitter designed to obtain three separate well trajectories from a common well bore.

Evidence suggests several attempts were made in California using such technologies.

In 1939, Leo Ranney drilled the first "horizontal well." To accomplish this, he drilled a large shaft, and then, placing men and equipment at the bottom of the shaft, drilled a horizontal section. He is also reported to have drilled horizontally in more than one direction, establishing not only the first horizontal, but likely the first multilateral with horizontal sections.

In 1953, Alexander Gregorian, a Russian drilling engineer, drilled what today would be known as the first truly multilateral or multibranch well (OGJ, July 6, 1998, p. 67). At that time, Russian engineers had acquired an extensive knowledge of downhole motors (turbodrills and electrodrills).

By using this technology along with minor steering capability, it became a natural progression to drill horizontal sections, and once horizontal, to drill open-hole sidetracks off the parent well (lateral) and branches off the laterals. Reportedly, production from these experiments was excellent, but for some reason, the technology did not catch on.

It was not until the late 1980s that substantial operations began in regards to horizontal and multilateral technologies (Fig. 1 [88,406 bytes]). For some time, it had been known that the Austin Chalk formations of South Texas were extensively fractured. When a vertical well intersected these fractures, it became quite productive. But, if the well missed the fracture sets, it came up dry.

Massive hydraulic fracturing procedures were performed on numerous wells to intersect the natural fractures with moderate success. Then companies like Oryx Energy Co., Union Pacific Resource Co., and others began to drill horizontally, crossing natural fracture sets with tremendous production results. Upon this successful application of technology, "the boom was on, land was quickly leased, and massive drilling campaigns begun."

This drilling boom is exactly what the industry needed to train personnel, develop equipment, and build confidence in horizontal wells. Multilateral wells became a natural spin-off.

The Austin Chalk, Buda, and other formations provided a natural testing. The consolidated, hard-to-damage formation allowed the wells to be left open hole, allowing almost any mud type to be used.

Multilateral analogy

A horizontal completion configuration is relatively straightforward. It is either open hole or cased. However, with multilateral configurations, there are many more options. In fact, the wide variety of multilateral configurations resembles the wide variety of automobiles available on the market.

There are many makes and models to choose from, yet very few people can name them all. Also like automobiles, multilaterals can be designed with specific purposes in mind. For example, some automobiles are used for transporting individuals to work and back home, while others are built for mass transportation across large distances.

Like automobiles, when people first go to the show room, they have something in mind. Once they find the price, they either scale down their expectations or keep their old vehicle. Multilaterals are like this in that it is necessary to have a clear expectation of what is needed, rather than what would be nice to have.

Multilateral definitions

To understand the technology, it is important to understand certain definitions (Fig. 3 [81,612 bytes]):

• A multilateral well is one in which multiple boreholes or laterals are drilled from a single well bore (Fig. 2 [38,148 bytes]). These boreholes may be horizontal or deviated in order to reach different bottom hole locations.
• Laterals are well bores drilled from the main well bore.
• Branches are well bores drilled from a horizontal lateral into the horizontal plane.
• Splays (fish hooks or herringbone) are well bores drilled from a horizontal lateral in the vertical plane.

In addition, there are a number of multilateral well trajectories to consider:

• A dual lateral is a multilateral well with two laterals.
• Opposed laterals are two laterals, opposed to each other at about 180?.
• Stacked laterals are two laterals drilled along the same azimuth at different depths.
• Junctions are the intersecting points from which laterals intersect with the main well bore or branches intersect with the lateral.

Finally, there are terms related to the equipment needed to complete a multilateral well:

• Whipstocks are devices with a hardened face used to deflect mills or the drillstring against the casing in order to create a new lateral.
• Deflectors are typically shorter than whipstocks and are primarily used to deflect tools from the main bore into the lateral or from the lateral to the branch.

There are two main categories of multilateral junctions:

1. A cased junction is one in which there is casing in the lateral that it is joined to the main well bore. This can be done mechanically through proper positioning and cementing, with expandable metals, or by drilling larger holes and installing premanufactured junctions (or junctions made on the surface).

2. The uncased junction is typically what the name implies. There is no casing in the lateral. The main well bore may be cased or not.

Historically, uncased junctions have been easier and less expensive to install. Therefore, uncased junctions are much more prevalent than cased junctions. However, to install an uncased junction, a suitable formation must be chosen that will not collapse. In many areas this is not possible and, therefore, cased junctions must be used instead.

Dogleg severity

By definition, to install a multilateral it is necessary to create a severe dogleg or abrupt bend at the junction. It is possible to control the dogleg severity (DLS) by directing either the whipstock angle or window length.

Typical whipstock tool-face angles are between 1.5 and 3°. This seems like a very minor angle, but increasing this angle from 1.5 to 3° will double the window height and almost logarithmically increase the DLS.

Fig. 4 [112,227 bytes] shows the relationship for 7-in. and 95/8-in. casing for these two angles. Also note that the DLS for drill pipe and casing in the same hole are different because of their relative placement differences.

There are numerous examples where the DLSs have grown too large. Even though it may be possible to pass drill pipe through the junction, it may not be possible to pass casing through the same zone. An increased DLS will also lead to an increase in torque and drag, and will reduce the burst strength of tubulars.

Junction depth

Multilateral junctions have been placed at almost any depth imaginable, from very shallow to very deep. However, junctions that are either not in the pay or of a Level 6 category (full pressure integrity achieved with main casing string), must be placed at a sufficient depth, or in a formation with sufficient strength, to contain the highest encountered reservoir pressure if a leak were to occur.

Fig. 5 [87,350 bytes] shows an example of how this depth may be determined using standard drilling pore prediction and fracture-gradient curves. It is important to add a safety factor to account for inconsistencies in predicted fracture-gradients and additional pressures that may be encountered during stimulation or kill operations.

Reasons to drill multilaterals

There are two basic reasons to drill multilaterals:

1. Mechanical or construction reasons-Typically associated with cost saving or slot saving reasons.

2. Reservoir or production reasons-Typically associated with improved drainage patterns or modified injection profiles.

To date, the industry has been driven by the construction push-save money, utilize the same well bore-rather than the reservoir engineering pull to improve flow patterns.

Typically, reservoirs that have good characteristics and from which accurate predictive models can be developed, will involve more multilateral, reservoir-driven well paths and completion scenarios that not only increase contact with the pay zone, but control flow from the reservoir. Reservoirs without such models use simpler multilateral configurations aimed at reducing costs.

Every reservoir has its own specific driver as to whether a multilateral configuration will be used. Reservoirs that are shallow, thick, and unconsolidated normally do not make good candidates for multilaterals. Many reservoirs that require sophisticated multilateral junctions can instead be developed more economically by drilling offset wells closely spaced from one another.

Drilling a multilateral well for the purpose of implementing a new technology is not a good reason for its use. In fact, if the plan is ill conceived and improperly implemented, it can do more damage than good for the technology.

To date, most multilateral laterals have been horizontal or at an angle greater than 80°. However, multilaterals can be of any well-path design. Before a company embarks on a horizontal multilateral or multibranch campaign, it should first become comfortable with horizontal technologies. There have been many failures where companies have tried to make "too big of a learning step too quickly."

There is a learning curve associated with multilateral technologies as there is with horizontal well technologies. In order to determine the true cost and test the application, a company typically needs to drill and complete three or more wells. Simply completing one well and sitting back to evaluate the technology will result in the loss of what little learning was previously attained, a well cost that becomes significantly greater than what can be achieved by drilling and completing additional wells. In addition, a one-well, wait-and-see strategy increases the probability of failure.

Before installing a multilateral system, it is necessary to ensure that the economics and reservoir development plans drive the use of this technology rather than just "demonstrating to management that you are using new technology."

Cost and risk

Cost and risk associated with multilateral well technology appear to go hand in hand. Most operators are reluctant to spend the equivalent cost associated with drilling two laterals as they would to drill separate wells.

Because multilateral technologies are still relatively new, before operators will take a chance with this technology, there either needs to be a significant cost savings or no other way to accomplish the same goal.

There are, however, several general trends in multilaterals that make them more appealing:

• The deeper the junction, the more attractive multilaterals typically become. For shallow wells or wells that drill extremely fast, it may be cheaper to drill two wells.
• The more wells drilled, the cheaper the technology. Never stake implementation of a new technology on one well. The cost will be more expensive and the lessons learned will be very hard to transfer, especially if only one well is drilled and evaluated for a long period of time before subsequent wells are drilled.
• The more laterals drilled from a well, the less the incremental cost for additional laterals will become. Learning curves improve and equipment, services, and personnel become fully utilized.
• Open hole branches (Level 1) are very easy to create and fast to implement, typically with no new equipment required.
• The faster the turning radius, the more the lateral will cost. Short-radius wells with many course corrections are typically more expensive than medium-radius wells.
• Separate wells are more expensive than commingled production. If the only reason you are separating wells is for data gathering, think seriously about this.
• Wells without a built-in reentry mechanism are less expensive than wells with built-in reentry capabilities. New coil-tubing tools and tractor technologies are improving, removing the need to build in the reentry capability along with the completion system. Also, as time passes, reentry systems by nature become prone to more failures.
• It is typically easier to install multilaterals by starting at the bottom of the well and adding junctions as the work progresses up the hole. However, many workers feel there is less risk of losing a lateral by starting at the top of the hole and working down. Although most companies complete from bottom up, both scenarios are successfully used.

Acknowledgments

The author will like to acknowledge Mobil E & P Technical Center for permission to present this article and Connie Woehr for help in preparing the manuscript and graphics.

Bibliography

1. Chambers, M., and Meehan, D.N., "Practical Issues in Multilateral Horizontal Well Completions," SPE paper 36455 presented at the SPE Technical Conference & Exhibition, Denver, Oct. 6-9, 1996.

2. Perdue, J., "Multilateral Technology Development Parallels That of Horizontal Wells," Sperry Sun Supplement, Hart's Petroleum Engineer International, 1997.

The Author