Drilling / Well Completion

How operators are benefiting from Level 3 multilaterals in Alaska

Level 3 multilateral completion technology is lowering costs and raising production rates. BP, Phillips and XTO have implemented development strategies based on the technology

Dave Westgard, Baker Oil Tools

espite its mature status, the North Slope of Alaska, with more than 15 billion bbl OOIP still untapped, remains an attractive prospect. The combination of unstable oil prices and drilling costs – typically three to five times those of drilling comparable wells in the lower 48 states or Canada – have dashed development hopes for some operators and seriously frustrated others. Now, however, BP has implemented a development strategy based on Level 3 multilateral completion technology, and other operators are following suit.

Since 2000, BP has completed six wells in the Milne Point Unit using Baker Oil Tools’ Level 3 HOOK Hanger proprietary system. Phillips Petroleum has completed two wells with the system in the West Sak sands in the Kuparuk area. Additionally, XTO Energy has used the system to complete a well in Cook Inlet.

Milne Point

The $179-million Milne Point, Schrader Bluff, S-pad expansion project, sanctioned in 2001 and currently under construction by BP, is the first large-scale North Slope development to be based entirely on the use of multilateral producers, Fig. 1.

Fig. 1. Location of Milne Point field.

The Milne Point Unit comprises three producing horizons: Sag River, Kuparuk and Schrader Bluff. (The latter formation is known as the West Sak in the Greater Kuparuk Area and the Schrader Bluff in the Milne Point Unit and Greater Prudhoe Bay.) Schrader Bluff is the shallowest of these horizons. It comprises three main sands – N, OA and OB – each separated by shale. The sands range from unconsolidated to weakly consolidated, and oil gravity is highly viscous, varying between 14° and 22° API. Sand permeability ranges from 10 to 300 mD in the O sands and from 200 to 2000 mD in the N sands.

Conoco began developing the Milne Point Unit in 1987, producing first from the Kuparuk sands, then installing four gravel pads in a pilot development area of the Schrader Bluff sands. After several years of study, Conoco deemed the project uneconomic. BP purchased 91% of Milne Point in 1994 and the remaining 9% in 2000.

Since field development began, electric submersible pumps (ESPs) have provided artificial lift throughout Milne Point for both Schrader Bluff and Kuparuk wells. Because excessive solids production causes ESPs to fail, and pump replacement cost averages $300,000, sand control was deemed necessary from the outset. With average ESP run life of 2.5 years, wells producing less than 200 bopd were considered uneconomic to repair.

Initial oil rates for the first nine years of the Milne Point production were 300 bopd per well. Throughout that time, Conoco, and later BP, searched for ways to increase well productivity while reducing drilling and completion costs. Conoco’s initial 17-well development focused on conventional well-completion technology with various sand-control techniques. The rationale was to increase oil rates by maximizing the number of sands completed in each well. One horizontal well was drilled.

From 1994 to 1996, BP field tested variations on frac-n-pack completions. In 1996, the company devised a large-scale Schrader Bluff development plan based on Frac for Sand Control (FSC) techniques that it believed would enable wells to produce up to 600 bopd. The $1-billion plan called for five new gravel pads, 10 mi of new roads and 75 mi of new pipelines. About 500 FSC wells would be drilled at a 7,500-ft radius from each pad.

The first phase of this plan was implemented in 1997, with a series of wells that were batch drilled in 1997, then batch completed in 1998. While 14 conventionally completed producers in this program produced a per-well average of 275 bopd, a non-sand control, horizontal stair-stepped producer came online at 600 bopd. Two additional non-sand control horizontal wells were drilled in 1998. Although their productivities were similar to previous conventionally completed wells, all three horizontal wells exhibited greater sand stability at higher rates than had been indicated by rock-mechanics studies and laboratory-sanding analyses. At the same time, a majority of the 1996 FSC wells experienced early and multiple ESP failures due to proppant and formation-sand production. At this point, large-scale field development was stopped and re-evaluated.

BP’s first multilaterals. The development team evaluated the use of long horizontal, multilateral wells, and reviewed the economics of completing only the weakly consolidated O sands and deferring development of the unconsolidated N sands. With horizontal well profiles enabling step-outs of over 10,000 ft from gravel pads, the new development concept showed that only one gravel pad (S-pad) would be required, with only one mile of new road and 7.5 mi of new pipeline. These parameters represented a dramatic reduction in environmental footprint, and significant cost reduction over the previous development scheme. Based on this concept, a new technical strategy was developed that focused on:

• Tripling well rates with dual-lateral sand wells
• Cutting operating costs in half by using jet pumps and long-life completions
• Reducing drilling costs 30% by using multilaterals and coiled-tubing drilling
• Reducing facility costs by increasing the reserve base.

The first multilateral wells in Milne Point were drilled in November 2000. Pre-drilling design considerations included: locating the junction (either in the sand or overlying shale); accommodating future lateral access; and planning for sand, water and/or pressure isolation at the junction. With no N sand in the well design and no need for sand control – the robustness of slotted liner completions had been proven earlier – the multilateral junction could be attained by a relatively simple Level 2 or Level 3 system. The development team considered future lateral access crucial to enable possible non-rig, coiled-tubing cleanouts. Additionally, the high cost of drilling in Alaska dictated systems that would minimize the number of trips and rig time.

The first three multilateral wells were drilled using Level 2 systems, with a world-first, hollow guidestock designed to allow coiled tubing re-entry into both laterals. The failure of a barefoot well from the 1999 program triggered a decision to modify the design. In 2001, the world’s first 4.5-in. x 7 in. HOOK Hanger Level 3 system was installed.

Hanger system. The hanger system provides mechanical support for junctions that join cased and cemented main bores with screened, open-hole laterals in wells with commingled production. In Milne Point, it allows the junction and the slotted liner to be set together in one run.

The introduction of this hanger system in January 1999 marked a significant step in the development of cost-effective, fit-for-purpose multilateral solutions, since it offered the option for re-entry into both the lateral and the main bore. Before this development, Level 3 systems provided little functionality since a slotted liner anchored back to the main bore limited re-entry to only the lateral. Achieving re-entry into both the lateral and the main bore required moving up to a higher-complexity, higher-cost multilateral system.

These hanger systems have simplified multilateral completion operations and have been used successfully in dual and tri-lateral applications. Depending on the application, the system can be run with either of two anchoring methods. Typical installation sequence for this hanger system is as follows:

1. The lower lateral is drilled and completed through the casing shoe.
2. A one-trip, proprietary, WindowMaster whipstock system with bottom-trip anchor is run in position to drill the upper zone and set on the liner hanger.
3. The casing window is cut, and the upper lateral is drilled to TD.
4. The whipstock is retrieved.
5. The Hook Hanger assembly is made up, run into casing and landed in the casing exit window. The system’s hook engages with the lower part of the casing exit window to hang the lateral-well system off the main-bore casing.
6. The well is now ready for final completion and production.

Retrievable lateral and main-bore diverters provide future access. Both can be deployed and retrieved with either coiled tubing or jointed pipe.

Currently in the Milne Point Unit, short step-outs are being drilled using standard North Slope drilling procedures. The more challenging extended-reach drilling (ERD) multilaterals have required special procedures to mitigate risks.

Current completion program. Solids production at Milne Point causes ESPs to fail on average every three years, adding an average $100,000 per year ($300,000 rig cost divided by three) to each well’s lifting costs. Conversely, jet pumps can produce solids without failure and require only slickline to change. The Schrader Bluff S-pad development project currently under construction will use jet pumps as its artificial lift system. The combination of through-tubing lateral access enabled by the Hook Hanger multilateral systems and jet pumps provides assurance that, if too much sand enters the slotted liners, they can be cleaned out without expensive rig workovers.

To date, the new multilaterals at Milne Point have averaged initial rates of 1,600 bopd, with one well producing 3,200 bopd in its first test. BP has now begun to experiment with coiled-tubing drilling to convert some of the earlier FSC completions into horizontal multilaterals, and thus double or triple their current production. The next phase of development will include incorporating unconsolidated N sands into the multilateral well design. Two wells are planned for 2003 to test the concept.

Phillips West Sak Multilaterals

Phillips Petroleum, operating in the West Sak formation in the Greater Kuparuk Area of the North Slope, had similar objectives to BP: lower current multilateral well costs by reducing wellbore trips; increase production from a mature field; and reduce multilateral risk. To date, Phillips has installed two of the hanger systems; one is described here.

When the completion team arrived at the drill site, the main bore had been drilled, and 7-5/8-in. casing had been run to depth and cemented in position. The lower lateral had been drilled to TD. The liner string for the lower lateral was picked up and run in the hole to the packer. The 2-7/8-in. inner string was then picked up and run in hole with proper cup space out. The liner string and the inner string were then run to depth, and the packer was set per procedure. The running tool was released from the packer, then the open hole and liner washes were completed. The sliding sleeve was closed and inner string was pulled out of the hole.

A new WindowMaster whipstock system was designed for Phillips to produce a straight, symmetrical, elongated window to accommodate the flanged-hook hanger. The new whipstock assembly was run in the hole, set at 19° left of high side. After milling, the upper lateral was drilled to depth and the whipstock retrieved.

The upper lateral-liner string was run in the hole to a Type B HOOK Hanger system, Fig 2. The inner string was assembled and run in the hole. The liner running tool, inner string and liner strings were assembled and run to a position above the casing exit. The bent joint was then oriented to engage the milled exit in the 7-5/8-in. casing. The upper lateral-liner string was run through the casing exit to a depth where the hanger was above the casing exit. The hanger was then oriented to the milled window using MWD correlation. Next, the hanger was landed in the window.

Fig. 2. Type B HOOK Hanger system, (top) is anchored with an external casing packer and includes a flange that fits around the casing exit window to minimize sand infiltration at the junction. This type has a larger window to the main bore than does Type S (middle), which uses slips to anchor the system in the main-bore casing.

The MWD mandrel was removed from the string using wireline. The running tool was released using a safety mechanical-torque feature. Finally, a no-go mandrel was run to verify position of the hanger. The no-go was passed through both the main bore and the lateral bore.

The simple installation method for this well reduced the number of trips by 50%, relative to some other multilateral systems, and saved over 100 hr of rig time. Estimated cost savings were $360,000. At the same time, production increased 280% over that of a conventional single well. Whereas single wells in this field have typically produced 200 to 300 bopd, production from the new multilateral wells has been 600 to 800 bopd.

XTO Installation at Cook Inlet

At 10,630 ft, this is the deepest HOOK Hanger installation to date; it was installed for XTO Energy in Cook Inlet. Multilateral operations for the well began by removing the existing completion equipment, including tubing string and seal assembly, down to the main-bore packer. After a scraper run to clean up restricted areas and verify the internal diameter of the 7-in., 32-lb/ft casing, the whipstock system was run in with MWD equipment included in the milling BHA.

The whipstock was oriented to about 16° right of the high side of the wellbore. It was then landed on the main-bore packer and the bottom trip anchor was set. Milling began at 10,616 ft and continued to 10,643 ft, including the full window and rat hole. The first milling assembly was then pulled from the well, and a second mill was run to ensure creation of a proper gauge window to allow the drilling assembly to exit. Drilling of the lateral well section was then started.

After the lateral leg was drilled, a clean-out run was made. The retrieval tool and MWD equipment were run to pull the whipstock, which was hooked at about 30° right of high side. After being pulled and recovered, the safety-shear disconnect was released and the whipstock was laid down on the surface. An overshot with a 3-in. grapple assembly and pump-out sub was run to pull the remaining whipstock anchor section.

The liner string, including a bent joint, slotted liner and the hanger system, was assembled and run in the well. The string did not require orientation as the bent joint entered on the first attempt into the lateral and the hanger was landed in the bottom of the casing window. The hanger was set with about 40,000 lb weight. A ball was dropped to release the hydraulic running tool. The ball seat was then sheared with 3,800 psi. The liner running string was pulled from the wellbore.

The wellbore was then conditioned and prepared for new completion equipment. The new completion string comprised an SB-3 packer, gas-lift mandrels and a TE-5 subsurface safety valve.

Previous single wells in this area had produced an average of 150 to 200 bopd. The new Level 3 multilateral is currently producing at 600 bopd. The hanger system has found a niche in this part of the world, helping operators significantly increase production while reducing well costs and risk.  

Acknowledgment

Much of the information about the Milne Point Unit in this article was taken from a paper, "North Slope Multilaterals Solve Problems and Increase Viscous Oil Production," by David P. Jenkins, Patrick J. Collins and Steve M. Schultz, all of BP, presented at Oil & Gas Journal’s Multilateral Well Conference, March 5 – 7, 2002, in Galveston, Texas. The author thanks Messrs. Jenkins, Collins and Schultz, BP, for permission to use this material. We also thank Phillips and XTO Energy for the opportunity to highlight their well.

The author
	Dave Westgard has been involved with multilateral and re-entry applications since joining Baker Oil Tools in 1997, where he is currently product line manager Multilateral Systems. His responsibilities include new systems development and marketing of multilateral systems worldwide. Previously, he spent 12 years in oil and gas drilling, which included conventional, horizontal and directional wells. He has instructed numerous classes and short courses related to multilateral wells and re-entry / sidetracking applications.