Case Studies - WA001 - Pneumatic Fracturing

Dry Media Application - Government Facility - Richland, Washington

Pneumatic fracturing was integrated with in situ vitrification (ISV) at an uncontaminated facility with a geology that consisted of a mixture of sand, gravel, and cobbles in Richland, Washington. The ISV process is designed to vitrify soil at intense temperatures, subsequently trapping contaminants within a glass matrix and volatilizing others. The project team devised an integrated Pneumatic Fracturing injection system to inject graphite particles into the subsurface to establish a conductive plane for the ISV process. This was the first field application of particle injection to be integrated with Pneumatic Fracturing.

Procedure

The project consisted of two parts: Phase I incorporated the analyses of soils, laboratory bench-scale testing, and system modeling and design. Phase II involved the field pilot demonstration at the site.

Outcome

Phase I conclusions:

The Hanford Formation was significantly different from other geologic formations which have been pneumatically fractured to date. Laboratory analysis of test samples showed its predominant textural components consisted of gravel and sand with a USCS classification of GP. It contained only 3.4 wt% of non-cohesive fines, and 0.12 wt% organic matter. The in-situ permeability of the formation was estimated to be in the range of 1,000 to 10,000 ft/day (0.352 to 3.52 cm/sec). Petrographic analysis showed that the soil particles were predominantly mafic igneous, with trace amounts of sedimentary particles.

Using the results of the soil analysis, a surrogate test soil was designed to model the Hanford test soil. Over 2,000 lb (907 kg) of the surrogate test soil was blended from local Northeast aggregate sources for use in the bench scale studies.

During the Phase I studies, a total of 35 fracture injection tests were performed in Plexiglass test tanks filled with the surrogate soil. The effects of several experimental parameters on the effectiveness of fracturing were studied. These parameters included moisture content, soil density, injection pressure, and injection flow. The tests were clearly subject to scale with similar tank tests performed over the last five years indicate that the bench scale tests were generally representative of mechanisms and critical parameters observed at the field scale.

During the tests, moisture content and soil density were determined to have a significant role on the ability to obtain a discrete fracture in the soil, and to subsequently inject dry media into the fracture.



The starter path media consisting of graphite and glass frit was successfully injected into a test tank to form a conductive link between two electrodes spaced 12 in (30.5 cm) apart. (Most injection testing was performed with plain air or a silica sand media to minimize contamination of surrogate soil). The conductive lens averaged 2 in (5.1 cm) in thickness, and conductivity tests showed an average resistance of 0.80 ohms across the electrodes.

Horizontal infiltrometer tests were conducted to examine the ability of the graphite and glass frit media to penetrate the unfractured soil matrix by interstitial transport rather than through fracture propagation. Experiments performed under a variety of conditions with different graphite gradations consistently showed minimal penetration of the conductive media into soil matrix at 0.13 in (3 mm) maximum in sand pores, and 0.2 in (5 mm) maximum at boundaries of oversize particles. These test results were verified with filtration and straining criteria for porous media published in the literature. Based upon both the laboratory and theoretical results, we did not expect interstitial transport to be a significant transport mechanism in this formation, except in locally coarse zones consisting entirely of cobbles and gravel.

The observed soil fracture mechanisms in the coarse-grained Hanford soils appeared somewhat different from the cohesive soils and rock normally treated with the pneumatic fracturing technology. Although the presence of moisture in these naturally non-cohesive soils would have provided some apparent cohesion, it was speculated that the formation would not have "fractured" in the normal brittle sense. Instead, the primary mechanism was better described as "pneumatic intrusion" or "pneumatic cutting." This behavior, coupled with the high leak-off of injected air into the unfractured parts of the formation, indicated that higher than normal flows and pressures would be required to treat this formation.

The ability to keep the graphite/glass frit media suspended in the injection stream and transport it in discrete, planar fractures was modeled using a few different theoretical approaches. The results indicated air velocities less than 0.3 ft/sec (<0.1 m/sec) which were sufficient to keep the mixture suspended. A radial injection model showed that leak-off into the formation of the injected air was considerable, and indicated that the radius of influence of the fractures would be significantly diminished compared with fine-grained soils and rock. The results suggested, however, that an effective radius of several feet would be possible, depending on the local formation texture.



Pilot Test area consisted of the large caisson (electrodes) surrounding the center injection pipe.



Dry Graphite injection system. Liquid nitrogen truck in background provided the injection fluid.

back