Extensive downhole measurements and heat flow surveys were obtained on both holes since they were originally drilled (Fig. 2). Despite the data collected to date which suggest that young, upper oceanic crust under a sediment cover is easily permeable enough to support considerable flow of seawater, the details of off-axis circulation and its control by the pressure distribution and fine-scale permeability structure are as yet poorly understood. Hole 395A is an excellent place for this experiment because returning to the natural thermal regime will allow determination of whether the observed downhole flow is dynamically maintained due to active circulation occurring in the basement, whether or not a borehole is present, or is due to flow induced by the geothermal gradient.
(1) the differential pressure (which should not be termed an "underpressure") between the cold hydrostatic pressure in a borehole drilled with cold seawater and the warmer hydrostatic formation pressure, and
(2) true, dynamically maintained underpressures due to active circulation in the basement that would occur whether or not a borehole were present.
In cases of downhole flow in holes drilled into formations with high geothermal gradients, e.g., the strong downhole flow observed during Leg 139 drilling at Middle Valley, the driving force is dominated by the former effect (Davis, Mottl et al., 1992). In the case of other holes that were drilled into crust with low geothermal gradients, e.g., Hole 395A, the latter effect may be predominant.
Hole 395A is one of the best-documented examples to date. The strong and consistent downhole flow requires both a significant pressure differential and long-lived formation permeability. The hole was sited where heat flow is very low, so there is very little differential between cold hydrostatic drilling fluids and formation temperatures to induce downhole flow. The fact that the downhole flow in Hole 395A has continued so strongly and for so long suggests that the dynamic pressure differentials due to an active circulation system are the predominant cause (and in fact these hydrologic processes may also be the cause of the low heat flow at the site).
Hole 395A is located in 7-m.y.-old crust in an isolated sediment pond (Hussong et al., 1979) that might be considered somewhat typical of the structure and hydrogeological setting for thinly sedimented crust formed at slow spreading rates. The hole has been revisited three times since it was drilled in 1975-1976 (Melson, Rabinowitz, et al., 1979): during DSDP Leg 78B in 1981 (Hyndman, Salisbury, et al., 1984), during ODP Leg 109 in 1986 (Bryan, Juteau, et al., 1988), and during the French wireline reentry campaign DIANAUT in 1990 (Gable et al., 1992). On each of these visits, the first order of business was temperature logging, because that measurement requires an undisturbed hole. Each of the three temperature logs showed strongly depressed borehole temperatures, essentially isothermal to a depth of about 300 m into basement (Fig. 2) (Becker et al., 1984; Kopietz et al., 1990; Gable et al., 1992). This indicates a strong downhole flow of ocean bottom water, at rates of thousands of liters/hr into the permeable upper oceanic crust, virtually unabated over the 20 years that the hole has been open (Fig. 3).
In comparison, temperatures measured during the multiple visits to Hole 504B were initially strongly depressed to a depth of about 100 m into basement, but then rebounded non monotonically towards a conductive profile, indicating that the rate of downhole flow has decayed since the hole was first drilled and that the downhole flow is directed into a more restricted section of upper basement (Becker et al., 1983, 1985, 1989; Gable et al., 1989; Dick et al., 1992). This comparison indicates that Hole 504B penetrates a more passive hydrothermal regime, while Hole 395A provides a man-made shunt into a significantly more active circulation system. The various observations at Site 395 generally support a model of lateral circulation in the upper basement beneath the sediment pond in which the site is located (Fig. 4) (Langseth et al., 1984, 1992), but we have little resolution on the details of such circulation.
The proposed program is intended to address these important issues by providing essential information about the formation pressure and permeability structure, which are real keys to understanding the crustal hydrogeology. Logging will begin with three logs designed first to provide a fourth estimate ofthe downhole flow at a single time point and to assess the fine scale distribution of permeability in the hole. The hole will then be CORKed with pressure sensor and thermistor cable, for a long-term record of the pressure and temperature variations in the sealed hole as the natural hydrologic system re-establishes itself. In more detail, we propose the following sequence of experiments:
(1) After initial reentry, a temperature log will be run followed by Formation MicroScanner (FMS) and flowmeter logs to delineate the fine-scale permeability structure of the section penetrated by Hole 395A. The upper 300 m of basement is known to be quite permeable on average from the downhole flow, packer measurements, and an incomplete flowmeter experiment during the DIANAUT program (Becker et al., 1984; Hickman et al., 1984; Becker, 1990; Kopietz et al., 1990; Gable et al., 1992; Morin et al., 1992). Detailed permeability information will be required to allow interpretation of the data collected from the CORKed hole (individual thermistor readings plus an integrated pressure) in terms of active hydrogeological processes in discrete zones of the formation. The FMS will provide a detailed log of fracturing (filled and unfilled), and a full flowmeter experiment will allow the integration of the fracture information with a detailed permeability distribution (given the strong downhole flow already occurring, the flowmeter experiment can be conducted without using a packer). This sequence would require a pipe trip, plus about 36 hours logging time, for a total of 48-60 hours.
(2) Deployment of a fully configured CORK to seal the hole, plus a thermistor string and pressure sensor in the sealed section (an actual mechanical latch to hold the CORK in the cone would not be necessary, as the differential pressure that has driven downhole flow will hold the seal in place). Such an installation would allow a long-term record of (a) the rebound of temperatures towards formation conditions after the emplacement of the seal, (b) possible temporal variations in temperatures due to lateral flow in discrete zones, and (c) pressure variations, which would be the primary manifestation in a sealed hole of changes in the forces that drive the natural circulation system. We would deploy a 600-m-long cable, with 10 thermistors spaced relatively equally below 100 m, to span the lowermost sediments, permeable extrusives and breccias, and the lowermost, relatively impermeable section of open hole (although the hole was originally drilled to 664 mbsf, during Legs 78B and 109 the deepest 55 m was found to be filled with cave-ins). The data logger and sensor string could probably be pulled out of the hole without the drillship, if necessary, but the CORK hardware would require the drillship for future removal. The deployment sequence would be similar to that required for deployment of the CORK already emplaced during Legs 139 and 146. A separate pipe trip would be required, plus about 6-8 hours to run in the cable and release the installation from the bottom of the pipe. Based on the Leg 139 experience, we estimate that a deployment in Hole 395A (nearly 2 km deeper than the Leg 139 sites) would probably require about 36 hours total.
The primary purpose of the CORK experiment would not necessarily be to assess the equilibrium pre-drilling thermal regime (which we can estimate from detailed heat flow surveys as in Fig. 2), but instead would be to monitor how the hydrologic system varies with time as natural hydrogeological conditions are re-established. Full thermal re-equilibration could require many tens or hundreds of years if it occurs by conductive processes only, but could also occur in much less time if the Langseth et al. (1984, 1992) model of active lateral circulation is correct. We are interested primarily in exploring the causes of the hydrogeological state and any possible temporal variations, with the simplest goal being to determine how these are associated with and controlled by formation pressure and/or permeability structure. It is impossible to model or predict all of the possible outcomes of the experiment, but considering two possible end-member results might be instructive:
(1) If the model of active lateral circulation is basically incorrect, and downhole flow is indeed simply an artifact of drilling, then sealing the hole should remove the driving force for the downhole flow, and temperatures and pressures will slowly and smoothly trend towards values consistent with conductive, hydrostatic processes.
(2) If there is some element of truth to the model of active lateral circulation in basement, with this circulation providing the driving pressure differential for the downhole flow, then sealing the hole will not change the driving force, and lateral circulation should continue even though the seal has stopped the downhole flow. Pressures in the sealed hole should approach a nonhydrostatic value in an irregular fashion that reflects variability in the natural hydrogeologic processes. Similarly, temperatures will rebound towards values consistent with the circulation system, also in an irregular fashion that reflects natural hydrogeologic variability. In addition, differences in the behavior of the temperature sensors should reflect vertical variations in the lateral flow regime due to fine-scale permeability variations. We understand so little about crustal hydrogeology that simply defining the natural time- and space-scales of such variability will be a very important result.
Hole 395A, drilled during DSDP Leg 45 (1975), together with 504B and other holes has been used for a long-term effort to determine the in situ petrophysical, geophysical, and geochemical properties of oceanic crust and toward the understanding of its formation and evolution. Comparative studies of in situ oceanic structures and hydrothermal circulation systems at oceanic ridges demand high quality and high resolution logging data. The logging data acquired before 1986 in Hole 395A at the Mid-Atlantic Ridge are poor compared with the high quality and high resolution logging data acquired in Hole 504B in 1993. In particular, the dual laterolog (DLL), critical to estimating porosity in high-resistivity formations, was not run in Hole 395A. A borehole televiewer (BHTV) downhole tool run was acquired by Morin et al. (1992) by wireline re-entry. With the recent advances in ODP logging capabilities, it is recommended that Hole 395A be logged again with the Quad combination with the array sonic tool, as well as the DLL, temperature, FMS, BHTV, and flowmeter tools. Time permitting with the limited number of logging runs, additional logs such as the dipole sonic and vertical seismic profile (VSP) would be extremely valuable for log integration with seismic data and for comparative studies of large-scale porosity, anisotropy, and hydrothermal properties in the ridge crest environment.
To Leg 174B-I Proposed Site Information