Recent Developments With Hydromechanical Well Tests

Hydromechanical well tests involve measuring and interpreting displacements along with hydraulic heads that result from a hydraulic stress applied to a well. The basic approach involves conducting a typical well test while deploying an extensometer, tiltmeter, or other device (e.g. InSAR) to measure displacements. The motivation behind this type of test is that the displacement measurements provide information about the constitutive properties and structure of the aquifer that go beyond what can be derived from pressure signals alone. The work we have conducted consists of slug and pumping tests with a moveable extensometer deployed between packers to measure axial displacements along isolated, 2-m-long sections of a borehole. Recent developments have included refinements in the instrumentation, evaluation of interference slug tests, and analyses in clastic sedimentary rocks. Displacements during well tests are remarkably small and the instrumentation we were using was limited to a resolution of +/- 150nm, which is too large to resolve signals in some of the fractures at our development site underlain by fractured biotite gneiss. We upgraded to a 24-bit analog-to- digital converter, and potted it in epoxy so it could be deployed next to the LVDT in the borehole to reduce temperature variations. This improved the raw signal to +/- 25nm and the resolution can be improved further using a low pass filter. The hydromechanical signal from slug tests was measured in monitoring wells and we confirmed opening displacements in monitoring wells 5 m or more from the source well. The most striking result was that the pressure in monitoring wells always decreased at the beginning of a slug-in test, only to increase soon thereafter as expected. This early water-level reversal lasted for 5s within 5 m, but it lasted for more than 100s in a monitoring well 10 m from the source well. This reverse water-level fluctuation during a slug test in fractured rock appears to be similar to the Noordbergum effect observed during pumping tests in sediments. Hydromechanical slug and pumping tests were conducted in fractured clastic sediments in the Newark Basin at a site near Trenton, NJ. The signals from the slug tests at one well appear similar to signals from tests conducted in fractured gneiss, except the compliance was remarkably stiff (less than 0.1 micron displacement/m of drawdown). Pumping tests at that well showed a flattening of the slope, likely the beginning of a dual porosity response. The fracture continued to close with no change in slope, however. Slug tests at another well consistently gave an anomalous response, with closing displacements occurring when the pressure increased during slug-in, and opening displacements when the pressure decreased during slug-out tests. The response was changed to the typical type of behavior by inflating a lower packer and isolating a fracture below the extensometer. Apparently a pressure increase in the lower fracture caused it to open and pinch closed the overlying fracture between the extensometer.