document

Abstract

In the spring of 1999, Burlington Resources initiated a well stimulation study in the Lewis Shale interval, San Juan Basin. The goal of the project was to establish the operational feasibility of pumping liquid CO₂ and sand (Dry Frac) hydraulic fracture treatments^1,2,3,4,5,6 in the Lewis Shale, and to compare production response of wells stimulated with liquid CO₂ and sand to those stimulated with aqueous based systems. Production comparisons and well tests were used to quantify the results of the Dry Frac technology.

The Lewis Shale is present across the San Juan Basin and is part of the Mesaverde formation at approximately 4000´. Of the wells stimulated in 1999, 16 were treated with liquid CO₂ and sand, a water free stimulation technology. The remainder of the wells (46) were treated with nitrogen-foam water based fluids. Prior work completed within the Lewis interval suggests that imbibing gelled fluid into the low-permeability, naturally fractured formation may cause permeability reductions. Permeability damage to the natural and hydraulic fractures may be eliminated and/or reduced by fracturing and propping the Lewis interval with the non-aqueous based fluids utilized in the Dry Frac process.

Introduction

From 1950 to 1990, only 16 wells were completed in the Lewis Shale across the basin. These wells encountered extensive natural fracture systems while drilling for deeper objectives. The wells were cased and completed naturally. The initial rates ranged from 1,000 to 10,000 Mscf/d with EURs ranging from 5 to 70 Bscf. Jennings⁷ summarized much of this information previously.

Between 1991 and 1997, Burlington Resources completed the Lewis Shale in over 110 existing and new wells. These were considered the more conventional Lewis Shale completions. Limited testing and/or reservoir characterization was done at this time. In 1998, Burlington Resources began their Lewis Shale program to determine the best portion to complete within the 1,000 to 1,500 ft interval, the appropriate stimulation method, and the expected reserves. Extensive work has been done to quantify the percentage of Lewis contribution in commingled production with spinner surveys. In addition, wells have been isolated with Lewis only production to quantify contribution from the interval.

Burlington Resources is currently collecting and analyzing extensive data in 10 Lewis “Shale Data Wells” to fulfill these objectives along with the spinner work and isolated production.⁸

To aid in quantifying what the appropriate stimulation technique should be, Burlington chose 16 wells to test the viability of liquid CO ₂ and sand treatments throughout the San Juan Basin. Conventional treatments use a nitrogen-foam stimulation in one or two stages. From post-treatment communication and well tests in the “Shale Data Wells”, Burlington has learned that during and after communication testing, substantial amounts of fluid have been produced into the wellbore. These tests were performed 5 to 7 months after the Lewis stimulations and the wells were on commingled production for this time period. Since the Lewis Shale formation does not typically produce water or any other fluids, further analysis was performed. It was determined by laboratory testing that the fluids contained treatment water, surfactants, and gelling agents, which were from the nitrogen-foam stimulations.

It is belivered that the stimulation fluid is imbibed in the natural fracture system and only a portion will be produced at typical delta pressure. If the producing pressure is reduced, as in the post-fracture testing, the imbibed water will move towards the wellbore. Typical stimulation fluid recovery is approvimately 30%. This leaves a substantial amount of fluid in the natural fracture system that can reduce the relative permeability to gas and hamper production.