Test Procedures

The procedure developed by Bjerrum & Andersen (1972), or modifications of the original procedure, is most often followed when conducting the HFT. The procedure was developed from in situ permeability tests using hydraulic piezometers that were pushed or driven into the ground. Tests are essentially conducted in the following sequence. The piezometer is installed to the test depth, and the excess pore water pressures generated during installation are allowed to dissipate to hydrostatic conditions. Once excess pore water pressure has

Figure 10.6 Test equipment for performing Hydraulic Fracture Tests.

dissipated, a screw pump is used to generate a constant head high enough to cause the soil to fracture. Pressure readings are taken during pumping to detect when the fracture has formed. Once fracture has occurred, the flow is stopped, and the dissipation of pore water pressure is monitored over time to determine when the fracture closes.

Using this procedure, it was postulated by Bjerrum &c Andersen (1972) that the minor principal total stress at that depth could be determined by monitoring the flow of fluid into the soil, which was assumed to be the horizontal principal stress across a vertical fracture. Modifications to the apparatus and procedure were made by Bozozuk (1974) by replacing the mercury manometer with a pore water pressure transducer and taking readings automatically with a chart recorder.

Results and Interpretation

To perform the test, the pressure is incrementally increased and the volume of fluid pumped is recorded, as shown in Figure 10.9. Fracture is noted by a dramatic drop in the pressure or by leveling off in the pressure as the fracture allows the fluid to enter. Pumping is usually stopped at this point, and the change in pressure is recorded. The dissipation of pore water pressure after fracture is analyzed by plotting the flow of fluid into the soil as a function of the decreasing pressure.

It is assumed that the fracture closes when a significant decrease in the flow into the soil is observed. The pressure at this point is usually taken as the total stress acting across the fracture or close-up stress, U_c. Bjerrum & Andersen (1972) concluded that U_c was equal

Figure 10.7 Use of a single packer to perform Hydraulic Fracture Tests.

to the minor principal total stress in the ground at the test depth, i.e., the horizontal total stress if a vertical fracture was formed. Bozozuk (1974) found that drawing two tangents through points prior to and following the break in the close-up curve gave good estimates of the minor principal stress, as illustrated in Figure 10.10.

After obtaining the close-up pressure following the initial fracture, it is often common practice to perform another fracture phase of the test. As shown in Figure 10.9, the pressure needed to reopen the fracture is also often obtained, typically indicated by the end of the linear pressure vs. volume curve. A second close-up curve is also sometimes obtained following the reopen part of the test. The complete sequence is shown in Figure 10.11.

Limitations to the HFT for determining lateral stresses in clays had been presented by Massarch et al. (1975), and Massarch & Broms (1976), as follows:

• 1. The uncertainty of the failure mode: varves, silt seams, fissures, and other nonuniformities may influence the failure mode;
• 2. The tensile strength of the clay is neglected in the interpretation of results;
• 3. The LIFT is limited to normally consolidated clays;
• 4. The shape of the piezometer (i.e., cylindrical) produces significant disturbance and may create arching during reconsolidation after installation. This may affect the measured valve in the test, giving preference to fracture at the piezometer tip rather than producing a vertical fracture.

Figure 10.8 Use of a double packer to perform Hydraulic Fracture Tests.

Based on the work performed since these limitations were considered, the author would offer the following comments:

• 1. The use of longer filter elements on the piezometer and L/D ratio greater than about 5 appears to produce a failure mode generating a vertical fracture. Discontinuities such as fissures are typically present in the most fine-grained deposits to some degree but are more predominant in surficial clay crusts, and therefore, some difficulty may still be encountered in these deposits.
• 2. Since the normal test interpretation is to use either the close-up or reopen pressure, neglecting the soil tensile strength appears to be justified.
• 3. The HFT is not limited to normally consolidated clays (K_o<1.0). Using longer filter tips positioned away from the piezometer tip has produced the reported K„ values of at least 3.0. Additionally, the author has experienced no difficulties installing push-in piezometers in very stiff overconsolidated clays and clay tills.
• 4. After complete reconsolidation, the value of horizontal stress measured by the HFT (when pushed) may be larger than that of in situ (before pushing) horizontal stress in

Figure 10.9 Typical results from Hydraulic Fracture Tests.

all but very soft normally consolidated clays. This is to be expected much in the way that piles driven into clays and allowed to reconsolidate will change the resulting stress field near the pile face. Therefore, the horizontal stress measured using HFTs with push-in piezometers should be referred to as the “effective hydraulic fracture pressure 0'_hc” and not Oh₀-

HFTs are reasonably simple and inexpensive to perform, and the interpretation is relatively straightforward. Results obtained indicated that K₍₎ values in the range of 0.5-3 may be determined. Table 10.2 gives a summary of some reported uses of HFTs.

Figure 10.10 Determination of close-up pressure by tangent intersection.

Figure 10.11 Sequence of Hydraulic Fracture Testing.

Table 10.2 Some reported results from Hydraulic Fracture Tests