1. Borehole-fluid samplers
Pristine formation fluids have never been retrieved successfully from a DSDP or ODP hole drilled into igneous basement or other hard formation. Recovering formation fluids from soft materials is relatively easy and routine during most ODP cruises: whole-round samples of cores are cut and squeezed in the shipboard laboratory. Hard (particularly igneous) rock does not yield pore fluids by squeezing, in part because hard rocks are difficult to squeeze but also because porosity in these materials tends to be fracture dominated and therefore not represented in recovered core. One means of obtaining samples of fluids from boreholes that penetrate igneous rock (fluid that may reflect the composition of pristine pore waters at some distance from the borehole) is to return to a reentry hole months or years after the last operations, after seawater pumped into the hole has had time to react with the surrounding rock, and gently lower a passive sampler into the open hole (Mottl and Gieskes, 1990). Eight such samples were obtained from DSDP/ODP Hole 504B at the start of Leg 137, using two customized samplers developed for geothermal applications.
Leakage is a common problem with sampling fluids in high-temperature wells. Once the hot sample has been collected and isolated in a chamber, the sampler must be drawn up through the cold-water column. Close to the seafloor, the sampler and fluid sample experience a rapid change in temperature with a relatively small change in hydrostatic pressure. Reductions in fluid specific volume associated with this change in temperature can create a vacuum inside the chamber, causing valves to leak, and allowing the sampler to take on additional (borehole or water-column) fluids. The intake of an unknown quantity of fluid of unknown composition can change drastically the chemistry of materials in the sampler (through mixing, reaction, dissolution, and precipitation). Designing leak-proof valves is probably the primary challenge in borehole-fluid sampling today.
The tools ODP used during Leg 137 were a mixture of established technology and developments by scientists and technicians working at Los Alamos National Laboratory (LANL) and the Lawrence Berkeley Laboratory (LBL). The tools were provided for this specific leg through special arrangement with the individual government laboratories, with ODP covering costs for shipping, maintenance, and technician time. Technicians from each lab sailed with the tools to handle preparation, maintenance, and sampling operations. Both tools employ a flow-through design, which allows flushing of the samplers as they are lowered through the water column (Fig. VIII-1). The design also prevents steam "flashing" of sampled fluids as they are drawn into an evacuated sample chamber, a problem with other sampling designs such as the WSTP. Titanium versions of both tools were available during Leg 137, hardened with special o-rings and other components to operate in temperatures greater than 300°C (Table VIII-1).
The two tools take different approaches toward overcoming the "shrinkage and leaking" problem associated with sampling high-temperature fluids: the LANL sampler has a collet (or notch) on each valve, intended to prevent it from moving significantly once the sampler is sealed; the LBL sampler includes an expandable-bellows pressure compensator. Both systems have advantages and disadvantages, and neither worked perfectly during Leg 137.
While the bellows in the LBL tool appears to have compensated sample shrinkage in several cases, it also restricted free flow through the tool, allowing it to entrain seawater as it was lowered into the hole. The LANL valves probably opened during tool ascent on at least one of the deeper runs in the hole, but this resulted in part because the upper colleted valve was incompatible with a screen needed to keep the valve free of rust, and so was not used. Another important difference between the tools is that the LANL sampler is a mechanical tool which runs on the coring line. The valves on the LBL tool are operated electrically, requiring a single or multiconducting cable. While the mechanical clock on the LANL has distinct advantages at high temperatures (where logging lines, connections, and terminations tend to fail), having the LBL tool run off a logging line also allows direct control over triggering time and therefore sample location. In contrast, there is no way to be positive as to when (and where) the LANL mechanism triggers. An additional problem with fluid sampling during Leg 137 was that of accessing the samples under pressure once they were returned to the surface. Extraction manifolds were provided along with the LANL and LBL tools, but these components were complex and could not be mated directly to shipboard analytical instruments in the chemistry laboratory.
An additional borehole-fluid sampling development is now under way. This effort is being led by P. Lysne (Sandia) and J. Edmund (MIT), who are working on the sampling apparatus, and K. Von Damm (New Hampshire), who is overseeing design and construction of a manifold and associated uphole plumbing for the tool. This tool is intended to be a high-temperature, slim-hole sampler, capable of running on a wireline and maintaining sample integrity during tool ascent. This tool is also intended to provide borehole temperature and pressure information. Later versions of this tool may also contain nuclear-logging capabilities.
The Geophysical Properties tool (Geoprops) was intended to provide information on the in-situ mechanical and hydrogeological properties of sediments in accretionary complexes. In 1987 D. Karig (Cornell) obtained NSF funding for a feasibility study of tool design. The 1988 feasibility study (authored by Karig and E. Taylor) described a tool comprising two inflatable, instrumented packers and a sampling apparatus, to be deployed through a conventional ODP coring bit into a previously drilled pilot hole. The pilot hole would be approximately 4 in. in diameter and drilled with a developmental coring system then being developed by ODP. Geoprops would provide information on formation temperature, pore pressure, permeability, in-situ horizontal stress, and pore-fluid composition (by sampling). Development and testing was intended to be completed in time for a scheduled leg to the Nankai accretionary complex in late 1989.
Funding for construction of Geoprops was obtained through NSF, and a contract was let with TAM International, the same company that had designed and built the successful ODP straddle packer. Unfortunately, TAM had little experience with the kind of development associated with Geoprops, and was also busy at the same time with design and construction of a wireline sampler for LDEO, so work on Geoprops was delayed. Problems also developed with the Navidrill coring system, which tended to stall; eventually this latter system was abandoned, and a new design was prepared for a motor-driven core barrel (MCDB) to replace the Navidrill. The Nankai leg was also delayed for 4 months, but the Geoprops probe was still not ready; Leg 131 to the Nankai accretionary complex sailed without Geoprops.
Two Geoprops tools were made available for land testing several months after the Nankai leg, in August 1990. Problems were soon discovered with the valve and shear-pin systems used to actuate different Geoprops functions, and some redesign was needed. A second series of bench tests was completed in June 1991, revealing additional valve, seal, and shear-pin problems. Additional recommendations from ODP engineers included redesign of the shock-absorbing system (necessary for intended free-fall deployment), compilation of electronics drawings and a list of spare parts, more extensive bench testing of individual components, and considerations of design upgrades affecting the sampling mechanism, materials of construction, and ease of operation. TAM felt they had fulfilled their obligation to the project and did not take part in these additional developments.
Karig decided not to pursue funding for these developments, but B. Carson (Lehigh) was supported by NSF to complete an additional phase of development and testing in time for ODP Leg 146 to the Cascadia margin. ODP engineer S. McGrath was assigned to oversee this new phase of development and testing, and a contract was let to an engineering consultant to do the actual work. Additional modifications and testing continued through the summer of 1992, and two modified Geoprops tools were sailed for "field testing" during Leg 146. Geoprops was finally deployed at Site 889 on the Cascadia margin, but the tool would not fully penetrate the MDCB pilot hole, so the packer could not be inflated. The tool was later dropped into the BHA and flow tested in the water column at Site 892 to learn more about its behavior under pressure. During this test, the Geoprops thermistors and shock sub failed. Additional modifications to the plumbing configuration and deployment method have also been recommended, although it is unclear at this time whether or not work with this tool will continue.
3. LAST I and II
Two generations of the lateral stress tool (LAST) have been developed for ODP use by K. Moran of the Atlantic Geoscience Centre, Geological Survey of Canada. LAST I is a passive coring device (Fig. VIII-2a) which is pressed into the sediments about 5 m ahead of the bit. The tool is configured similar to a standard APC coring shoe, except that the shoe's taper is on the inside rather than the outside of the cutting edge. While this disturbs the sampled core more than regular piston coring, it also presents the formation with the flat cylindrical face of the outside of the shoe. The tool is left in the formation for about 30 minutes, during which time strain is measured with three gauges mounted in the wall of the tool. The gauges are oriented 120° apart in a single horizontal plane. Pore pressure and temperature at the wall of the tool are also measured. The strain gauges are bonded to the inside of thinned sections of the tool wall, which act as diaphragms that differentially deflect under lateral load. The outside of the diaphragm is intended to measure total lateral lithostatic stress, while the inside of the diaphragm experiences ambient pore pressure. Thus the deflection of each diaphragm should reflect the effective lateral stress. Like the APC temperature tool described in Chapter III, LAST I is entirely self-contained and programmed with an IBM-compatible PC. Measurements are made at user-defined time intervals, typically every 5 seconds while the tool is in the formation. The tool is powered by six AA batteries that are changed for each run. LAST I was first fielded at sea during ODP Leg 131, and was run again during Leg 146.
LAST II is an active measurement device intended to determine in-situ lateral stress, strength, elastic modulus, and pore-pressure response (Fig. VIII-2b). The measurement element is a radially expanding bladder with two pressure sensors. One sensor measures expansion pressure, and the other measures formation pore pressure. The bladder is expanded using a motor-driven piston. Piston displacement is calibrated with bladder expansion to determine radial displacement of the bladder.
The tool is deployed with the same colleted delivery system used for the WSTP (Chapters II and III). The tool is pushed into the formation about 2.6 m ahead of the bit, and left in place during operation. Lateral stress is determined from the lift-off pressure of the stress-strain curve. Shear modulus and strength are determined directly from the stress-strain curve. Pore pressure is measured independently. The tool is programmed using a bench-top computer, then disconnected and sent down the pipe. The tool cycles through its test sequence via downhole computer control. LAST II can be programmed for several loading and unloading cycles; the tool was first fielded at sea during Leg 146.
LAST I apparently functioned properly during two of five initial deployments on ODP Leg 131, although there were minor glitches related primarily to shipboard protocols and tool programming (Moran et al., 1993). Data from LAST I and LAST II runs during Leg 146 have yet to be analyzed to determine if they are scientifically useful.
4. Flowmeter Experiment and Logging Cable Go-Devil
As described in Chapter IV, measurements of bulk permeability in the upper oceanic crust have been conducted successfully using drill-string packers in the DSDP and ODP. These measurements reveal the equivalent "bulk" or average Darcian permeability of the zone between an inflated packer and the bottom of open hole. However, permeability in the oceanic basement and other settings such as accretionary complexes is likely to be dominated by fractures and irregular networks of inter-connected porosity, as suggested by many logs and the discrepancies between bulk permeabilities measured downhole and laboratory permeabilities of competent samples. The flowmeter experiment was intended to assess the detailed permeability structure in a borehole, using constant-rate fluid injection into a zone sealed by the drill-string packer, with concurrent logging of borehole pressure and flow velocity in the isolated zone. Thus, this experiment involves both a logging tool and the drill-string packer, and it required careful coordination among the third-party investigators, LDEO wireline personnel, and ODP engineers.
An NSF grant to K. Becker (Miami) and R. Morin (Miami and USGS) provided funding for the design and construction of the flowmeter, a spinner tool made by Comprobe. In addition, ODP engineer T. Pettigrew designed a new packer go-devil to accomodate the experiment, and this go-devil may have application to any logging tool intended to be run in a section of hole sealed by the drill-string packer. Before the development of this go-devil, logging or other experiments were not permitted through an inflated drill-string packer, because this would have involved unsafe procedures on the rig floor, as follows: When the packer is inflated and gripping the borehole wall, the BHA is immobilized and the drill-string heave compensator must be engaged to avoid tearing the packer loose. With the drill-string compensated, the upper end of the pipe will move relative to the heaving rig floor, and it would normally be very difficult and unsafe to attempt to "stab" a logging tool into the end of the pipe before running it down into the isolated section of hole.
The new go-devil is initially attached to the logging cable above the cablehead connected to the flowmeter (or other) tool, and the flowmeter, logging cable, and new go-devil are deployed before the packer is inflated and the drill-string heave compensator is engaged. When the go-devil lands in the packer, it enables packer inflation in the standard way. After the packer is inflated and the drill-string heave compensator is engaged, additional pumping then allows the logging line to detach from the go-devil, and the flowmeter can be lowered into the isolated section of the borehole (Fig. VIII-3). This approach allows the logging tool to be introduced into the isolated zone before the drill string becomes heave compensated, and thus avoids the potential safety problems described above. Although there are presently no plans to run other logging or sampling tools in combination with the drill-string packer, in principle it should now be possible to do so.
Funding for the flowmeter was granted about one year prior to its first intended use on Leg 137. Comprobe was able to build the tool in time for land-based field testing prior to the leg at the Palisades sill test well, with the assistance of LDEO personnel. The flowmeter experiment was not successful during Leg 137, because the packer lost inflation pressure and inadvertently deflated. However, flowmeter experiments were successfully completed during Leg 139, as described in Becker et al. (in press).
5. "Active" fluid sampling
The WSTP and Kuster tools are described as "passive" samplers (Mottl et al., 1983, 1985) in that the fluids sampled are those that are in the immediate vicinity of the sampler, whether in sediments penetrated by a probe or in open hole. In some situations it may be more desirable to actively draw a sample from the formation, particularly in semilithified or lithified sediments that cannot be penetrated by the WSTP probe, or in open basement holes where the borehole fluids have not had time to equilibrate with pore fluids. During Leg 111, a Schlumberger repeat formation tester (RFT) was run in Hole 504B as the primary sampling tool to extract a pore-fluid sample from the wall of the borehole. The RFT is a "doughnut" sampler run on the logging line, with a circular sealing port that is pressed against the side of the borehole using a mechanical arm. The RFT failed to obtain a formation sample, returning borehole fluid instead, and the mechanical sampling arm broke off in the hole. Subsequent analysis of RFT sampling technology suggests that this design will not return a scientifically useful sample from a hole through a fractured formation if the borehole wall does not include a mud cake (common in commercial petroleum wells), as fluid drawn into the RFT will simply bypass the rubber seal from the open hole (R. Desbrandes, pers. comm., 1992).
The Borehole Research Group at LDEO later undertook to develop and test a wireline sampler for active pore-fluid sampling during ODP. The original intent was to produce an instrument that could be deployed through the pipe and bit over any zone of interest, like a conventional logging tool. The tool was designed and built by TAM International, the same company that produced the successful ODP straddle packer described in Chapter IV. Wireline-sampler development extended over several years, and the tool was finally deployed during Leg 133.
The wireline sampler consisted of a short straddle packer, four sample bottles, downhole pumps and valves to inflate the packers or pump fluids from the formation to the sample bottles, and electronic sensors to monitor key properties of the fluids to be sampled, such as pressure, temperature, pH, and concentrations of Na and Ca. The tool was intended to be deployed in open hole, with elements that would inflate from the 3.9-in. minimum internal diameter of a regular BHA to 10-13-in. in the borehole. The tool would begin a pumping sequence, with real-time surface readout of fluid properties to allow the operator to assess when some component of "pristine" pore fluids was being pumped past the sensors. The operator would then trap one to four samples of fluid, either in time series at a single level, to allow extrapolation of pure pore-fluid composition, or at several different levels within the hole during the same deployment.
Unfortunately, the main filter assembly became clogged and collapsed under pressure during deployment in Hole 816C on Leg 133 (Davies, McKenzie, Palmer-Julson, et al., 1991). In addition, the inflatable packer elements did not return to a small enough diameter upon deflation to pass back up through the BHA. Analysis of the wireline packer design after Leg 133 led those involved in its development to recommend numerous design modifications, requiring a substantial investment of time and money. As it was not clear that it would be physically possible to obtain a scientifically useful sample from a hard formation disturbed by drilling, the wireline packer project was tabled pending the outcome of the "Request for Proposals" for assessment of active fluid- sampling technologies, as described below.
Without a riser system (which would allow the use of heavy mud to avoid circulation loss) and a long-term, stable well-head facility (which would allow well "production" for days or weeks), it may not be possible to draw formation fluids from the wall of an ODP borehole that are undisturbed enough to be of scientific use. In some ways, the borehole CORK was intended to address this need. The CORK isolates a borehole and allows the borehole fluid to equilibrate with the surrounding formation; unfortunately, the CORK also requires considerable time, money, and effort for installation. Also, reacted borehole fluids are not exactly the same as formation pore fluids.
In response to requests from the community, PCOM in 1992 convened a Fluid-sampling Working Group to produce a Request for Proposals (RFP) for a feasibility study of available and developmental technologies for pore-fluid sampling in hard rock and lithified sediments. The RFP also requests an assessment of the hydrogeological state around an ODP borehole to determine under what conditions there is a good chance for successful fluid sampling. PCOM has deferred distribution of this RFP, due to a lack of available funds, until the Spring of 1993. Depending on the results of this RFP, and the resulting feasibility study (which could indicate that formation sampling in hard rock is beyond the reach of ODP operations), some form of active sampling technology may become available in the next 2-3 years.
6. Additional Developments
Additional tools are presently available or are being developed through LDEO for upcoming legs. These tools include: slimline (Diamond-coring compatible) natural gamma-temperature-caliper and Kuster mechanical temperature tools; a high-temperature, digital borehole televiewer; a slimhole, digital borehole televiewer; a multichannel, shear sonic tool; and high-temperature resistivity and sonic tools. Another third-party development presently underway is a slimline, pressure-temperature-flowmeter-conductivity-caliper tool. Contact the science operator or logging contractor for up-to-date information on availability and specifications of these instruments.