Fluid circulation in the oceanic crust and through the seafloor occurs in passive continental margins, accretionary and nonaccretionary prisms, mid-ocean ridges, and ridge flanks. The placement of constraints on the physical and chemical nature of crustal fluid flux can be done by direct observation, measurements on cored material, and by downhole measurements and sampling. Downhole measurements and sampling have proven to be the best method for deep, high-pressure environments. In areas where the natural flux is large, numerous Deep Sea Drilling Project and ODP holes have been drilled. Unfortunately, flow into the formations being drilled is induced by nonhydrostatic formation conditions and/or the differential pressures resulting from density differences between the warm, low-density formation fluid and the cold, higher-density drilling fluid in these boreholes. CORK installations are designed to prevent this flow and to monitor in situ temperature and pressure conditions as drilling disturbances decay and natural conditions return. Borehole fluid samples representing formation fluids over extended time intervals can also be obtained without requiring CORK removal (Davis et al., 1992). CORKs have been deployed in ODP holes during Legs 139 (Juan de Fuca Ridge), 156 (Barbados accretionary prism), 168 (Juan de Fuca Ridge), 174B (Mid-Atlantic Ridge), and 195 (South Chamorro Seamount) to further understanding of fluid budgets. The CORK deployed in Hole 1200C will be the first to permit study of a nonaccretionary prism environment.
Total fluid budget determination at accretionary active margins has been hindered by the presence of laterally heterogeneous and transient flow processes, with the former resulting from different flow rates and compositions along the strike of the margin and the latter resulting largely from the valvelike influence of the sediments of the accretionary complex. Because the hydraulic flow properties of sediments vary with fluid pressure and fluid pressure varies as a function of fluid production rate and transient hydrologic properties, the accretionary system acts in a nonlinear transient manner as both a seal and a relief valve on the fluid flow system.
A total fluid budget should be more readily determined at nonaccretionary active margins because the heterogeneous and transient effects from the sediments of the accretionary prism do not have to be considered, resulting in hydrologic flow systems that operate on longer timescales, approaching steady-state flow. This hypothesis must be tested by determining the physical nature of fluid flow in nonaccretionary settings. Fluid budgets can then be constructed to determine whether the expected long-term flow is consistent with observations or if the flow is transient.
The design used for the CORK in Hole 1200C incorporated most of the features used in previously CORKed holes, as documented by Davis et al. (1992). These features include the following:
The components used to meet the above objectives were a data logger, two pressure transducers, a thermistor cable, an extension cable, a spectra rope, two osmotically pumped fluid samplers, a weak point, and a sinker bar (Fig. F69; Table T19). These components were joined and, in some cases, fabricated on board the JOIDES Resolution while the hole was being drilled. The spectra rope, thermistor cable, and extension were joined with electrical tape, tie wraps, and wax string at ~5-m intervals to prevent the tape from unraveling and the tie wraps from sliding down the cable as the rope stretched under tension. It was important that the rope, thermistor cable, and extension cable be permitted to stretch independently, since each had a different stretch coefficient. The measured pH of 12.5 was of some concern because no one knew how the various cable components would withstand 1.5 yr in this corrosive environment.
The data logging unit is contained in a 4130 alloy mild steel pressure case that provides the upper CORK seal when latched into the landing collar (Fig. F70). Because the landing collar is constructed of the same material as the pressure case, the possibility of corrosion is minimized. The instrument hardware is mounted on an aluminum chassis within the 50-mm (2.5 in) inside diameter (ID) pressure case. Padded bulkheads and soft mounts at the top and bottom of the chassis provide longitudinal shock absorption.
Analog to digital conversion and 16 channels of voltage and resistance may be recorded by the data logger at intervals of 10 s or greater. Data are stored in 2 MB of solid-state memory, permitting data accumulation for a period of up to 5 yr with 1-hr sampling. For Hole 1200C, 15 channels, set for 1-hr sampling, were utilized: date and time, borehole pressure, seafloor pressure, internal temperature, resistances of the nine thermistors of the thermistor string, and two fixed, low temperature-coefficient resistors. Power is provided by four lithium thionyl chloride "C" cells, which will run the instrument for up to 5 yr under typical operating conditions. Additional backup power for memory is provided by a separate lithium cell.
Data retrieval and programming commands are performed over a 9600-baud three-wire RS-232 link. The external data communication link is provided by a four-contact, coaxial underwater mateable connector mounted on the top of the data logger (Ocean Design Inc.; "O.D. Blue") (Fig. F70). Power can be supplied to the data logger via this connection from an external source. The electrical connection can be established with a specially designed, gravity-driven, self-centering "top- hat" mounted on a submersible or ROV.
Temperature-compensated quartz pressure transducers with a total range of 40 MPa (at 4000 mbsl) were employed to measure the seafloor and borehole fluid pressures (Paroscientific Model 8B 4000-2) (Davis et al., 1992).
The thermistor cable string consists of 164 m of 12 twisted pairs of polypropylene insulated No. 20 American wire gage (AWG) wires with a nominal wall thickness of 0.016 in wrapped in an outer armor of Kevlar. Breakouts for the thermistors are underwater connectors (SeaCon MAW2-HC). The lower end of the cable is sealed to prevent fluid seepage into the data logger electronics.
This cable was originally deployed in Hole 1024C in 1996 during ODP Leg 168 prior to being recovered by the Atlantis (Woods Hole Oceanographic Institution) in September 1999. Because the hole temperatures were low (~30°C), because recovery resulted in no cable damage (aside from cutting off the T1 breakout), and because subsequent testing indicated good insulation and conductivity, the cable was considered to be in good condition for use in Hole 1200C. Because it was too short for the projected hole depth, extensions were fabricated to enable emplacement of thermistors at the desired depths while permitting the use of the existing breakouts. Because of hole problems, the actual hole depth obviated the need for the extensions.
The thermistors (Thermometrics SP100) were molded into MAW2-HC connectors. The thermistor cable had breakouts at depths of 29, 64, 99, 134, 139, 144, 149, 154, and 159 m of the paired conductors molded to 1N2 cables, which were in turn molded to MAW2-HC connectors. The breakouts were taped with 2-in electrical tape, with care taken to place the molded connections and thermistors against the cable body to prevent thermistor damage. Hydraulic hose (1.25 in ID) protected the installation, which was secured with electrical tape, nylon cord, and wax string (to prevent the tape from unraveling).
The spectra rope had been used during earlier CORK recoveries and was of suitable length for this deployment and design. This rope was used to make two separate lines with galvanized thimbles at each end. Rope 1 (149 m long) was fastened to the bottom of the data logger, tied off at each osmotic sampler, and terminated at the weak point. Rope 2 (25 m long) was fastened to the other end of the weak point and terminated at the sinker bar (Table T20; Fig. F69). The addition of a weak point was a change implemented to address the possibility of hole fill preventing sampler recovery. In Hole 1200C, both osmotic samplers were deployed above the screen, inside casing, with the weak point located at the top of the screen and the intake for the lower sampler located inside the screened section of casing (Fig. F69). The weak point consisted of a galvanized shackle (galvanization was removed to increase rate of corrosion) with an aluminum pin whose design had been tested to failure between 2 and 2.8 T. With no corrosion protection, it is thought that the weak point will most likely break prior to the recovery attempt.
Developed by Hans Jannasch (Monterey Bay Aquarium Research Institute) to provide a long-term nonelectronic sampler without moving parts, osmotic samplers can continuously collect borehole fluids for up to 5 yr. Constructed primarily of polyvinyl chloride and Teflon tubing, they are driven by molecular diffusion of water through a rigid semipermeable membrane that separates a saturated salt solution from distilled water. The resulting osmotic pressure is used to pull water through one membrane at a rate of ~4 µL/hr at 20°C. Because the rate of flow increases with increasing temperature, the precise rate of flow is determined by the number of membranes and the operating temperature. Osmotic pumps draw from a distilled water reservoir, which is connected to a small-bore Teflon tube (1.1 mm ID) that is open at the sampling port. The length of the tubing is determined by the pump rate and deployment time (Davis, Fisher, Firth, et al., 1997). Although diffusion tends to integrate samples with different compositions, the characteristic distance for molecular diffusion should be <0.5 m for the expected 1.5- to 2-yr duration in Hole 1200C.
Each osmotic sampler in Hole 1200C has five membranes, with the operating temperature to be determined from the thermistor string data. Because the two samplers will be deployed for 1.5 to 2 yr (recovery is expected in spring 2003), each one had 4000 ft of Teflon tubing. The top sampler (osmotic sampler 2) sampling port is located at a depth of 142 mbsf and the lower sampler (osmotic sampler 1) has a 1.1-mm ID sampling tube that extends below the weak point into the screened section 3 m above the sinker bar to a depth of 171 mbsf (Fig. F69).
The samplers (7.62 cm outer diameter) pass through the CORK and over the spectra cable (Fig. F69). The deepest thermistor is located just above osmotic sampler 2 so that temperature, and thus rate of pumping, will be known for the life of the experiment. Samples will be analyzed for the major and some minor seawater ions. The stable isotopic compositions of H, O, C, B, and Cl will also be measured.
Deployment was conducted in eight phases. In order, they were as follows:
An overview of the deployment methodology as stated by Davis, Mottl, Fisher, et al. (1992) is found below. A description of the first three steps can be found in "Hole 1200C" in "Operations." A detailed description of the last five steps can be obtained from the ODP Drilling Services Department. Figure F71 provides a visual representation of the last five steps, and Figure F70 shows the final result. Figures F72 and F73 provide insight on the drill string equipment used.
Six sections of 5.5-in drill pipe (58 m), referred to as the stinger, were attached to the bottom of the CORK prior to deployment. The stinger is used to keep the CORK centered in the reentry cone during deployment of the thermistor string/data logger. It also serves to add weight (8500 lb, in this case) to the string to prevent heave from prematurely unlatching the running tool from the CORK. The CORK running tool, which, "jays," or latches onto the CORK, was attached to the end of the drill string (Fig. F72). Once the running tool was "jayed" onto the CORK, a hydraulic latch-setting hose was made up between the running tool and the CORK.
The CORK was then lowered to the seafloor on the drill string. Once the reentry cone was located, the CORK was positioned with the stinger in the cone throat such that the CORK could not land or latch in the cone as the ship heaved (Fig. F71A). Because the coring line used to deploy the thermistor string/data logger was not heave compensated, the drill string was held in this position while the thermistor string/data logger assembly was deployed.
The data logger, thermistor string, osmotic fluid samplers, and sinker bar were placed in the drill pipe and attached to the logger seating tool (Fig. F73A), and the assembly was lowered down the drill string with the coring line. The sinker bar and thermistor string passed through the CORK, the stinger, and into the cased and screened borehole below. The data logger was then seated and latched inside the CORK (Fig. F73B). After the data logger latch-in was confirmed, the data logger seating tool was unlatched from the data logger and retrieved (Fig. F73C, F73D).
Once the wireline was out of the drill pipe, a CORK-setting go-devil was pumped down the drill string (Fig. F71C). The heave compensator was activated, and the drill string was lowered until the CORK with logger assembly landed in the 10.75-in casing hanger. The go-devil landed in the CORK running tool and blocked circulation, allowing the drill string to become pressurized. The pressure was channeled to a latch-setting piston via the hydraulic setting hose. When activated, the setting piston drove a latch ring into mating grooves in the 10.75-in casing hanger (Fig. F71D). The CORK must be latched in place to prevent it from being pumped out of the reentry cone should a positive pressure differential be present in the borehole.
After verifying that the CORK was latched in, a submersible ROV platform was deployed. Consisting of a central sleeve and a horizontal plate with holes that covers the reentry cone, it allows a vehicle to rest stably beside the CORK. The platform is fabricated in two halves. Prior to deployment, the halves were positioned around the drill string and bolted and welded together. Using the drill string as a guide, the platform was then free-fall deployed, automatically centering itself over the CORK as it landed on the reentry cone rim (Fig. F71E).
After verification by camera that the platform was in place, the running tool was "unjayed" from the CORK by setting down, rotating the drill string clockwise, and then lifting up. Removal of the running tool from the CORK resulted in the hydraulic setting hose automatically disconnecting. Thirteen hours after deployment began, the drill string was recovered and the instrumented borehole seal installation was complete (Figs. F69, F70, F71E; Table T21).
CORK data logger downloads can be performed by a submersible or ROV. Also, because CORKs have two internal hydraulically controlled systems (a differential pressure vent and data logger latch), hydraulic power supplied by a submersible or ROV permits removal and replacement of the thermistor string/data logger without requiring the capabilities and expense of a drill ship. The procedure is (1) a wireline is connected to the data logger, (2) the vent and the latch are hydraulically opened in sequence, and (3) the data logger/thermistor string is removed using the attached wireline. By reversing the unlatching sequence, replacement of the data logger or other instrumentation can be performed (Davis, Mottl, Fisher, et al., 1992).
An ROV data download for Hole 1200C is tentatively scheduled for spring 2003. It is hoped that funding and ship time can be obtained to recover the osmotic samplers, thermistor string, and data logger and redeploy the data logger during that visit.
Only a drill ship is capable of recovering the main body of a CORK through the use of a pulling tool attached to the end of the drill string. Lowering the pulling tool over the CORK mechanically activates the differential pressure vent to open, and the pulling tool "jays" onto the CORK release sleeve. Lifting the drill string engages the CORK release sleeve, which releases the CORK latch ring. The release sleeve then engages the CORK itself, which is free to be pulled out of the reentry cone.
The CORK and the submersible/ROV platform is then picked up until it is sufficiently clear of the reentry cone to prevent heave from reseating it. Because only the stinger (the drill pipe attached below the CORK) is positioned in the reentry cone throat and casing at this time, the heave compensation can be deactivated and wireline equipped with a mechanical latch can be lowered down the drill string to retrieve the thermistor string and data logger. Once this assembly is retrieved and the CORK with the submersible/ROV platform is recovered with the drill string, the borehole is open for further operations (Davis, Mottl, Fisher, et al., 1992).
Prior to drilling the CORK hole, single-bit exploratory Holes 1200A and 1200B were drilled to determine what procedures were needed to successfully drill the CORK borehole. Because Holes 1200A and 1200B collapsed, as anticipated, they were not cemented. We believe that the low hydraulic conductivity of the formations will prevent communication between these holes and the CORK without grouting these exploratory holes. Drilled to a depth of 145 mbsf, Hole 1200A was the deepest and poses the greatest threat of communication with Hole 1200C, the CORK hole. The distance between Holes 1200A and 1200C is ~100 m, and Hole 1200C has a total depth of 266 mbsf, with the screened section between 149 and 203 mbsf and no bridge plug present. Hole 1200B, before collapse, had a total depth of 98 mbsf and was located near Hole 1200A. The distances between Holes 1200D, 1200E, and 1200F from Hole 1200C are 52, 137, and 111 m, respectively (Fig. F3). Because each of these holes penetrated <30 mbsf, none should pose a risk to Hole 1200C.