The Bakken formation in the Williston basin has experienced an explosion of activity since the year 2000 through the effective use of technology and the quest to optimize operations and

Few horizontal wells had been drilled in the Elm Coulee area by 2000. (All graphics courtesy of IHS Inc.)
reduce costs. As a result, IHS Well and Production Data reports production rates in the Elm Coulee area have increased from approximately 10,000 bbl of oil per month in 2000, to rates exceeding 1.5 million bbl of oil per month today.

Horizontal well completions are now yielding economically attractive initial-production (IP) rates averaging 425 b/d oil (although some IPs have been reported prior to fracture stimulation) and recoverable reserves averaging more than 500,000 bbl of oil per well, according to the “Middle Bakken Member (Lower Mississippian/Upper Devonian) Richland County, Mont.” paper presented at the June 11 to 13, 2006 American Association of Petroleum Geologists Rocky Mountain Section meeting in Billings, Mont.

Various horizontal well bottomhole configurations have been drilled with up to seven laterals being completed from a single well bore. Combining multilaterals with new completion technology has resulted in highly productive wells, and has made the Bakken formation one of the most active plays in the United States.

The area included in this analysis is the Elm Coulee area in Richland County, Mont., which is located on the western side of the Williston Basin. The reservoir is the Bakken formation with a stratigraphic trap developed in the Middle Bakken member, according to Society of Petroleum Geologists paper 90697, “Improved Horizontal Well Stimulation in the Bakken Formation, Williston Basin, Mont.,” by Wiley, et al.

The middle member of the Bakken provides up to 14 ft (4.3 m) of reservoir, which bears up to 12% porosity, has a permeability of less than 0.2 mD and oil saturations in the 75% range. The reservoir is also slightly over-pressured. Local variations in reservoir properties and fracture development have been a major factor in well productivity throughout the Elm Coulee area.

Elm Coulee area history
The first commercial Bakken well at Elm Coulee was completed in 1981 by Coastal Oil and Gas. Early wells were vertical completions and were often a fall-back position for an unsuccessful deeper well-test. After stimulation, vertical Bakken wells have historically exhibited a respectable initial production, but with high initial declines followed by long-term, low productivity. A typical economic vertical well would initially produce at more than 100 b/d of oil, but would soon decline to approximately 15 b/d to 20 b/d of oil.

Production thereafter exhibited a minimal decline and due to the low production rates, oil price was the major factor in economic operations. Gas/oil ratios were approximately 750 Mcf/bbl, however, no sales outlet was available, and casing-head gas was used on the lease.
By 2007, more than 800 laterals had been drilled in 500 Elm Coulee area wells.
Superior productivity was dependent upon intersecting local fractures to double or triple the IP and the corresponding long-term productivity rates. The estimated ultimate recovery (EUR) of the typical economic vertical Bakken completion is estimated at 130,000 bbl of oil. Assuming a 30-year project life, however, that results in unfavorable project economics.
The first horizontal, Upper Bakken wells were drilled in Richland County in 1989 and 1990, but few were successful. The EUR rates for these horizontal wells were estimated at 21,000 boe. In 2000, Lyco Energy initiated the current horizontal activity by drilling the Burning Tree State 36-2H using an improved completion technique. The Lyco well’s IP was 196 b/d of oil, 85 Mcf/d of gas and 7 b/d of water, and the EUR is currently estimated at more than 300,000 bbl of oil.

Recent activity
In the last seven years, more than 480 wells involving 730 laterals have been completed and were reporting production in March 2007. The total number of horizontal Bakken wells drilled in the Elm Coulee area is now more than 500 and includes more than 800 laterals. This increase in activity in the Elm Coulee area since the first of 2000 versus year-to-date 2007 is shown in Figures 1 and 2.

Completion techniques
The first Bakken vertical wells were fractured using standard techniques and water or hydrocarbon-based fluids and sand. Initial-production rates were often encouraging, however, the vertical wells exhibited high initial-decline rates and the low-matrix permeability provided long-term, but uneconomic production rates.

Beginning in 2000, horizontal well-completion methods used cemented lateral liners and limited-entry perforation techniques in an attempt to control fracturing. Efforts were made to control vertical-fracture growth into the overlying Lodgepole or the underlying Three Forks (reservoir seals). Surveys indicated the lateral toe was preferentially being treated using this technique.

Modifications have since been made to the completion technique which include, but have not been limited to, non-cemented and pre-perforated liners; longer laterals; cleaner fracturing fluids; staged treatments with diverters; more aggressive gel breakers; and increased proppant volumes. These modifications have made a significant improvement in well productivity and also have resulted in a cost reduction.

Operators have since experimented with the configuration of the laterals. Different multilateral configurations have been drilled to find the ideal number of laterals to achieve optimum productivity, improve recovery from non-core areas and to reduce costs. A history of the evolution of wellbore configurations in the Elm Coulee areas is summarized in Table 1.

The question to answer
By analyzing performance of the 440 horizontal wells included in this study, the author seeks to help answer a question that the operators and investors would like to understand, which is “What is the wellbore configuration that should be drilled to yield the optimum economic return in this formation?”

Evaluation methodology
In order to evaluate the individual wellbore configurations, wells were grouped by the total number of laterals drilled from each vertical well bore. Production history was summarized for each type of wellbore configuration, and a normalized average well was created for each configuration.

Production forecasts were then made based on each normalized average production history and taking into consideration the long-term performance of vertical wells drilled in the 1980 and 1990s. Economics were developed for each normalized average well based on the economic parameters in Table 2.

Assumptions
For this evaluation, the following assumptions were made:
1. Completion technology has progressed equally across all configurations.
2. A homogeneous reservoir with no variation between wells.
3. Lateral orientation is identical for all wells, but could well be a needed component.

Results
Economic evaluations of each typical well were made using PowerTools a decline-curve and economics analysis software tool from IHS. The results are shown in Table 3.

Summary
Based on the economic indicator of discounted (10%) cash flow and the assumptions previously stated, single laterals or dual laterals are projected to be the most economically effective in the Bakken formation. Investors and operators should be aware that single- and dual-lateral wells could well be a better option than wells with more laterals. Unless there are other major factors to consider, careful analysis of investment in wells with more than two laterals should be done.

Next Steps
There could be well be additional factors to consider when making the decision to drill a multiple wellbore configuration Bakken well. Factors such as surface access limitations, anticipated lower-quality reservoir, company reserve volumes, etc., were not considered in this evaluation.
Additional investigation should focus on several topics including, but not limited to:
1. Investigating the effect of geographical location.
2. Investigating the effect of lateral orientation.
3. Investigating the effect of fracturing rates and pressures.