Previous   |   Next

doi:10.2204/iodp.proc.301.106.2005

Packer experiments

Hole U1301A

The drill string packer was deployed in single-element mode in casing in Hole U1301A to assess the permeability of upper basement between the base of casing (14.9 m into basement) and the bottom of open hole. Inflation of the packer within the open hole was precluded by poor hole conditions and the large diameter of the hole, which was drilled with a 14¾ inch bit. The packer was positioned at 267 mbsf, 10 m above the casing shoe (Fig. F70), and the ship's mud pump was used for all operations. The experiments were conducted after the pipe trip to drill out the cement at the 10¾ inch casing shoe and a check of open hole depth without rotation or circulation (see "Operations"). The depth check had indicated an obstruction at a depth of 34 m into basement, compared to total drilled depth of 107.5 m into basement. Given that the depth check was made without rotation or circulation, we assume that the obstruction was incomplete and that the hydraulically tested interval comprises the 92.6 m section between casing and total drilled depth. However, an alternate interpretation is also possible: that the only interval open to the hydrologic testing was the 18.7 m between casing and the depth reached during the depth check.

Two Geophysical Research Corporation Model ERPG-300 electronic pressure gauges (SN 78743 and 78745) were deployed in the packer go-devil. Their records of pressures and internal gauge temperatures are shown in Figure F71. After the go-devil landed but before the packer was inflated, hydrostatic baseline pressure was recorded for 1 h. Then the packer was inflated at ~1000 psi (6.9 MPa), after which the sealed-hole pressure baseline was recorded for 1 h. A series of five constant-rate injection tests was conducted at 15 spm (equivalent to 5 gpm or 0.3 L/s) for 30 min, 30 spm for 45 min, 50 spm for 1 h, 75 spm for 2 h, and 100 spm for 1 h. Following each period of injection, the pressure recovery was monitored for a period of the same duration as the respective pumping time. The last three injection tests were of longer duration than any attempted previously in DSDP or ODP and were intended to investigate to greater radial distance into the formation than previous packer experiments in upper oceanic basement.

The pressure records from the two gauges (Fig. F71), which had been recalibrated by the manufacturer just before Expedition 301, are offset by ~15 kPa because the depths of the sensing points differed by ~1.5 m. During periods of no pumping, the records show strong trends in the baseline readings (pressures increasing with time) that we interpret as transient recovery effects from the thermal perturbations to near-borehole fluid densities due to circulation during prior drilling operations. Owing to the timing of these operations and the depth of penetration of the hole into an overpressured setting, these trends are stronger than any observed in previous packer experiments in other crustal holes. They are also opposite in sign to the additional transient thermal perturbations induced by the pumping of cool fluids into the formation during the injection tests themselves. Because of the masking effects of these opposing thermal effects, the pressure rises recorded during the injection tests do not conform to behavior commonly seen during injection testing (in the limit, linear with log time). Thus, full interpretation of these tests will require correction to the recorded pressures for the two kinds of thermal perturbations to borehole fluid densities and pressures, calculations that may require careful numerical modeling.

During the injection tests, the magnitudes of the responses recorded by gauge 78745 were greater than those recorded by gauge 78743, despite the fact that gauge 78745 was installed in the shallower, somewhat more isolated position in the go-devil. The difference in gauge response cannot be explained easily, but may be partly a result of hydrodynamic effects during the relatively high volume pumping combined with relatively low signal level in terms of response from the highly permeable formation. Although the uncorrected pressure data do not conform to ideal behavior, preliminary shipboard interpretations for all but the first injection test at the lowest rate were made using Glover's formula (Snow, 1968), which applies to the case in which the pressure rise on injection reaches a nearly steady state value. We used the abrupt pressure drop at the end of injection as an approximation to the nearly steady state pressure rise achieved during injection. The results (using the signals from either gauge) indicate a permeability on the order of 3 × 10^–11 to 5 × 10^–11 m² when averaged over the 92.6 m basement section drilled below casing. This preliminary value lies between indications of (1) lower permeability at shorter length scales from shorter-duration Leg 168 packer experiments in Holes 1026B and 1027C and (2) higher permeability at much larger scales on the same ridge flank from earlier CORK observations and thermal modeling. Thus, preliminary indications are that the results are consistent with and complement previous reports of a spatial scale effect of permeability of uppermost basement on the flank of the Juan de Fuca Ridge (Becker and Fisher, 2000; Becker and Davis, 2003).

Hole U1301B

In Hole U1301B, the drill string packer was successfully inflated in single-element mode at three positions in open hole that were chosen primarily on the basis of caliper and density information from the wireline logs. In sequence, the three packer seats were at 472 mbsf (3140 mbrf), 442 mbsf (3110 mbrf), and 417 mbsf (3085 mbrf). In contrast to Hole U1301A, conditions in Hole U1301B were excellent (aside from the gap in the 10¾ inch casing string above the tested interval), and the hole was open to the full drilled depth of 582.6 mbsf, or 317.6 m into basement. The inflation/test sequence was designed to determine the permeability of upper basement at deeper levels than tested in Hole U1301A, for one of the few assessments of the depth variation of upper basement permeability in the ocean crust. More specifically, the test sequence was conducted to allow (1) direct measurement of the bulk permeabilities of three zones of upper basement between those inflation depths and the bottom of open hole and (2) indirect assessment of the permeabilities of the thinner zones between inflation depths by taking differences of transmissivities calculated for the larger zones measured directly. At the conclusion of the hydrologic testing program, the packer was also successfully inflated shallower in the hole for a brief test of the suitability of a likely seat for the CORK packers at 3047 mbrf, or 379 mbsf.

The same ERPG-300 electronic pressure gauges used earlier were deployed in the packer go-devil, and the ship's mud pump was used for all operations. The records of gauge pressure and internal temperature for 78745 are shown in Figure F72. After the go-devil landed at the first inflation depth but before the packer was inflated, hydrostatic baseline pressure was recorded for 1 h. Then the packer was inflated at ~1000 psi (6.9 MPa), after which the sealed-hole pressure baseline was recorded for 1 h. At the first inflation depth, a single slug test was conducted, but this decayed rapidly and only constant-rate injection tests were conducted thereafter at all inflation depths. Under the expectation that permeabilities deeper in basement in Hole U1301B would be less than those in the shallower basement section in Hole U1301A, lower rates of injection were used in Hole U1301B: 11, 20, and 30 spm for the inflation at 472 mbsf and 15 and 30 spm for the inflations at 442 and 417 mbsf. Injection periods were all 1 h, and each was followed by a 1 h period of pressure recovery.

As in Hole U1301A, the pressure records from the two recently recalibrated gauges were offset by ~15 kPa because the depths of the sensing points differed by ~1.5 m. Also as in Hole U1301A, the magnitudes of responses recorded by the two gauges differed significantly, despite the relative consistency of the hydrostatic reference readings. When the gauges were recovered from the testing in Hole U1301B, the screened carrier for the gauge showing lower-magnitude response (78743) was observed to be fouled with rust particles, and it is inferred that this may have attenuated the gauge response. This carrier also held the same gauge for the testing in Hole U1301A, during which that gauge also recorded a lower-magnitude relative response. Thus, the record from gauge 78745 is considered a more reliable indicator of actual borehole pressure response during the testing in both Holes U1301A and U1301B.

During periods of no pumping, some of the records show trends in the baseline readings like those seen in Hole U1301A (pressures increasing with time) that we interpret as transient recovery effects from the thermal perturbations to near-borehole fluid densities due to circulation during prior drilling operations. However, the effect was less significant in Hole U1301B than in Hole U1301A, with the exception of the intermediate inflation depth of 442 mbsf. Again, full interpretation of these tests will require correction to the recorded pressures for the thermal perturbations to borehole fluid densities and pressures, calculations that may require careful numerical modeling.

Nevertheless, the uncorrected pressure data conform to ideal behavior much better in Hole U1301B, partly because the thermal perturbation effects were less and also because lower injection rates were used in Hole U1301B. As for Hole U1301A, preliminary shipboard interpretations were made using Glover's formula (Snow, 1968), using the nearly instantaneous pressure drops at the ends of injection periods as approximations to the near steady-state pressure rises achieved during injection. The results suggest permeabilities on the order of 1 × 10^–11 m² for the three zones tested, spanning 152–317.6 m into basement. This preliminary value is less than the 3 × 10^–11 to 5 × 10^–11 m² suggested for the shallower basement interval in Hole U1301A. If borne out by more complete analysis, the preliminary processing would suggest only a slight reduction of permeability of upper basement with depth at Site U1301 and a thicker zone of highly permeable uppermost basement than interpreted at the older ridge flank sites (ODP Sites 504 and 896) in 6 Ma crust on the Costa Rica Rift (Becker, 1996).

Top of page   |   Previous   |   Next