A NEW SAMPLING SYSTEM FOR OBTAINING RELATIVELY UNDISTURBED SAMPLES OF UNCONSOLIDATED COARSE SAND AND GRAVEL

Geologic site characterization is an important issue that must be addressed if details of contaminant transport in the subsurface are to be understood and predicted. Ideally, the formation heterogeneities at a site must be characterized at several scales. This paper deals with the smallest practical scale: core samples with cross-sectional dimensions of a few inches. Historically, there has been great difficulty in obtaining relatively undisturbed cores of unconsolidated coarse sand and gravel. In this study, 23 wells were installed at a field site in the Kansas River alluvium with hollow-stem auger techniques. Six of these wells were cored through about 30 feet of coarse sand and gravel at depths of 40 to 70 feet using various techniques. A modified Waterloo sampler was employed with good success, using drilling mud in the auger flights to control heave and help hold in the sample. However, the use of heavy drilling mud has disadvantages (for example, the potential to contaminate the aquifer and cores) and recovery without sample loss is difficult because the procedure is sensitive to vibration and other mechanical forces. Without the use of drilling mud, the modified Waterloo sampler design was unsatisfactory because of relatively low recovery. In order to address this limitation, new sampler designs were developed and field tested. The most promising design does not require drilling mud, achieves a high recovery, and is not as sensitive to vibration and mechanical forces during recovery. The new design incorporates an inflatable bladder, located in the drive shoe, which closes off the end of the sampler. The deflated rubber bladder lies behind a plastic sample liner as the core begins to enter the sampler. Near the end of the 5-foot sample drive, a piston extension triggers a release mechanism and allows a 4-inch retraction of the plastic liner, resulting in the bladder being in direct contact with the sediment. The bladder is then inflated from the surface with nitrogen gas, closing off the bottom of the sampler and allowing recovery with minimal opportunity for sediments to fall out. Using this sampler, we have been able to consistently achieve about 86 percent out of a possible 90 percent recovery (drive shoe loss is 0.5 feet out of 5 feet due to bladder length and placement). The remaining 4 percent loss is due to compaction, premature piston movement, or wall friction preventing material movement into the sampler. These cores have been evaluated in the laboratory for determination of hydraulic conductivity, porosity, density, and particle-size fraction.