Previous   |   Next

doi:10.2204/iodp.proc.301.104.2005

Submersible operations

Table T3 summarizes operations with manned and unmanned submersibles at ODP-era CORKs after their initial installation. The level of submersible activity at the CORKs has been extensive, involving many of the international deepwater research vehicles. The costs for this activity have been supported primarily by national ODP funding agencies outside of commingled ODP funding. From the first dive with Alvin in 1991 at the first CORKs, the submersible pilots and support engineers for all the vehicles we have used have met nearly all of the needs and challenges associated with operations at the ODP CORKs. Submersible operations at CORKs have typically included routine data downloads at periodic intervals, sampling or pumping operations at the wellhead valve, and occasional complex instrument recoveries and reinstrumentation attempts; further details on these three aspects are provided in this section.

Data downloads and logger reprogramming have been accomplished by incorporating a single UMC in each CORK, with the mating connector linked for serial communication to a personal computer via through-hull wiring into the submersible, to the surface ship via remotely operated vehicles (ROVs), or, in two instances, to the surface ship via two-way acoustic modem (Meldrum et al., 1998). For the original CORK, the UMC was the economical "OD Blue" connector made by Ocean Design, Inc. The male connector incorporated four bands for contacts along a single pin that required no azimuthal orientation for mating, and it was assembled into a "top hat" unit that fit over the top of the CORK wellhead and assured proper alignment for mating with the female connector centered on the top of the data logger (fig. 2 of Davis et al., 1992). This system worked well aside from some quality control problems with the spacing among the banded contacts, such that full contact was not made with certain male units in certain female units. Careful selection of the connectors actually used prevented any malfunctions in the field. Manufacture of the OD Blue model was discontinued at about the same time that the multizone CORK models were being designed; these later installations utilized more expensive, better quality, but more difficult to mate multipin CM-2008 series ROV connectors made by Branter/SeaCon.

Hydraulic coupling to the wellhead valves has been required for access to sealed hole intervals, whether for sampling fluids, pumping into zones for active hydrological tests, or calibrating the pressure records. Most CORK installations have utilized a standard Aeroquip (model FD76-1002-08-10) hydraulic quick-connect/quick-disconnect system that is commonly used on farm tractors. The coupler is generally incorporated into a mating device with a T-handle so that the submersible pilot can push it into place; occasionally, there have been problems making the connection when the valve position was awkward and/or the submersible had difficulties overcoming the reactive forces during the coupling process. Tubing of the desired material is run off the coupler to the sampling device or pump on the wellhead or in the submersible basket. This system has proven satisfactory where sealed-hole pressures are superhydrostatic but unsatisfactory where the formation is underpressured. Efforts are being made to improve the submersible coupler to guarantee a tight seal under conditions of negative or low positive formation pressure.

For active pump-testing of zones isolated in CORKs, the pumps (supplied by scientists to date) must be carefully designed to interface with submersible electrical and hydraulic systems. The designs may vary depending on whether the objective is high pump rates in relatively permeable formations, yielding low pressure signals, or low pump rates, yielding high pressure signals in tight formations. In one unfortunate example of the latter (Hole 949C), local pump-generated transients working against the low formation permeability exceeded the working range of the pressure gauge, which then failed completely. Thus, design refinements are probably in order for the submersible-based pump systems for any active formation testing done at CORKs during IODP. In new installations, such as the ODP ACORK and CORK-II designs, such problems can be avoided by using three-way valves that have been incorporated. These allow the gauges to be switched to ambient seawater and isolated from any excessive transients during pumping, and they also provide a simple way to perform hydrostatic checks for intergauge calibration and drift tests.

At eight single-seal CORK installations, attempts have been made to recover original data loggers and sensor strings some years after initial deployment, either to retrieve downhole OsmoSamplers or because the instrumentation had been damaged, and most of these holes have also been resealed with a dummy plug or reinstrumented with a data logger with pressure sensors but no sensor string. No attempt has been made yet to deploy a long replacement sensor string using submersibles. Various methods have been used to recover the sensor strings, all first requiring the submersible to install a tool at the wellhead that latches onto the data logger and retracts the mechanical "dogs" that had originally latched the logger into the CORK body. In some cases, the latching tool has been attached to a rope run to flotation on the sea surface; after recovery of the submersible, the rope was pulled onto the ship and the packages were winched out. In one case (Hole 948D), the latching device was attached to flotation that had been predeployed to the seafloor with counterweights; the instrument string was floated up when the counterweights were released. In other cases, a bail was attached to the latching tool so that the MPL control vehicle could engage it and pull the instrumentation using the main ship's winch. In another case, an ROV was used to install the latching device, which was attached by strong line to the ROV "garage": once the ROV was retracted into the garage, the ship's winch was used to recover both the ROV system and the sensor string. Finally, the OsmoSampler packages for the Costa Rica margin CORK-II installations were run on landing plugs that seated deep in the hole. A submersible-mounted winch system was used in the attempts to recover those strings, but without success. Recovery and replacement of those strings required the drillship early in IODP operations.

It should be noted that in two attempts to recover sensor strings with OsmoSamplers (Holes 1025C and 1026B), the basaltic formation had closed in on the portion of the sensor string in open hole; during the unsuccessful recovery attempts, significant pull (up to ~5000 lb) had to be generated to break the thermistor strings, leaving the OsmoSamplers in the holes. In an earlier installation without OsmoSamplers (Hole 892B), hydrates had apparently frozen in around the sensor string, even within casing, and again significant pull (~7000 lb) had to be applied to break that string, in an attempt to clear the top of the hole for other experiments.

Reinstrumentation has generally involved predeploying the data logger with a running tool attached to a system of weights and floats that allows the submersible first to release predeployment weights to maneuver the package into the hole under near-neutral buoyancy and then to release the floats and attached running tool for the return to the surface. The recovery and reinstrumentation operations have utilized unlatching and running tools designed by ODP engineers. Many of these tools are currently warehoused at Texas A&M University (College Station, Texas, USA) and presumably could be made available for similar operations in the future.

Top of page   |   Previous   |   Next