Previous   |   Next

doi:10.2204/iodp.proc.301.104.2005

A few scientific lessons learned and challenges for the future

Each of the series of original and multilevel CORK installations completed to date has benefited from the experience gained from the preceding deployments. Technical examples are discussed above, and in this section, we focus on more basic issues critical to the scientific return from sealed-hole observatories. In some cases, the time between programs was insufficient to allow proper scientific and engineering evaluation and response (e.g., between Legs 196 and 205), but in most instances, modifications could be made to account for previously unanticipated problems, to study processes in ways previously unrecognized, and to incorporate new technology. For example, it became evident from the first deployments (ODP Legs 139 and 146) that a local record of seafloor pressure was necessary to account for and properly understand the formation pressure response to seafloor loading. Astronomically derived tidal constituents calibrated with satellite data and tide-gauge records could not be relied upon at the precision required, and no information could be had about other loading constituents across the broad frequency range of interest. Hence, subsequent installations included seafloor pressure sensors identical to those used for monitoring formation pressure. The paired data have allowed seafloor loading (as a source of noise) to be removed from the formation records and have provided novel constraints for estimating elastic and hydraulic properties of the formation (with seafloor loading as a formation signal source). We include here a brief list of other lessons and challenges that we believe are important for planning future CORK installations. These should be considered in the context of common first-order goals for CORK instrumentation, namely

• To obtain a high-fidelity record of pressure that has minimum phase or amplitude distortion over as broad a frequency range as possible,

• To record as accurately as possible formation temperatures as a function of depth and time, and

• To permit collection of representative formation-fluid samples that are as free from the effects of drilling and postdrilling contamination as possible.

The list is by no means comprehensive, but it should serve to help guide future designs.

• Correct observations of both transient and average pressure state depend critically on the quality of the seal created by the CORK system. Leakage through the CORK seals or plumbing, between the casing and formation, around packers, or through nearby unsealed pilot boreholes can cause pressure losses. Pressures in formations with high permeabilities and storativities seem to be relatively insensitive to minor amounts of leakage, although high rates of associated flow can cause thermal perturbations and associated pressure offsets. If the thermal perturbations are well constrained, the latter can be accounted for, but they undermine the confidence with which interpretations can be made. Pressures measured in low-permeability formations are sensitive to leaks in a more direct manner; large pressure offsets can be created with little flow, and thermal perturbations may not provide a good test for leakage. In all settings, great effort must be made to create leak-free installations and thermal observations should be made to test for flow and to constrain the buoyancy perturbation if flow is thermally significant.

• Another source of pressure signal distortion is system compliance. This can arise from compressibility of the fluid filling the cased boreholes in original CORKs or the umbilical tubing in ACORK or CORK-II installations and from compliance of the tubing or of the packers or seals that isolate monitored levels. The compliance of the thick-walled steel hydraulic tubing used to date is negligible, but at high frequencies, the compressibility of the water filling the umbilicals can couple with the high resistance of low-permeability material to filter high-frequency signals. Care must be given to minimizing the diameter of hydraulic tubing and to purging lines of any air or free gas. Little is known about the role of packers, but they may cause problems with high-frequency signal distortion in low-permeability formations.

• Combined fluid sampling with temperature and pressure monitoring must be approached with care for the reason that the means of sampling can constitute a leak. In most instances to date, fluid sampling has been done with samplers sealed into the formation, so this has not been a problem. In the multilevel CORKs installed during ODP Legs 196 and 205 and Expedition 301, provision was made for sampling at the seafloor via small-diameter umbilical lines. With this configuration, a proper balance must be achieved that allows a rate of flow that is great enough to make the transit time from the sampled level to the seafloor acceptably short but not so great as to cause a loss in pressure or a distortion of the thermal structure. As in the case of real leaks, this poses the greatest challenge in low-permeability formations.

• Great value has been realized in very long, continuous records. Of particular interest have been transients related to seismogenic strain. These signals are relatively rare and often small and can be characterized only in the absence of installation transients and with the careful removal of seafloor loading effects. Their frequency content is very broadband, ranging from short-lived (seismic surface waves, instantaneous elastic strain) to quasi-permanent deformation. Changes in the character of the response to loading have also been observed at the time of such events. Studies of such phenomena require uninterrupted records that are influenced minimally by such things as changes in sensors (resulting in calibration offsets and changes in drift) and by perturbations associated with hole opening.

• Perturbations caused by the invasion of cold, high-density seawater into the formation during drilling and any time after when holes are unsealed can be very large, particularly in high-permeability formations. The effect is exaggerated in subhydrostatic holes. Such unnatural flow affects temperatures and pressures and displaces formation water, precluding the collection of pristine formation fluid samples. Choosing sites that are naturally superhydrostatic helps to overcome this problem, as does minimizing the time between when permeable formations are first penetrated and when they are sealed. An additional problem with original CORKs arose with the large volume of water contained in the cased sections below the seafloor seal. This amounts to an unwanted reservoir of seawater trapped after installation that slowly mixes into the formation. This problem created challenges for the sampling efforts in the Leg 168 holes (Wheat et al., 2003). The multilevel CORKs designed for Legs 196 and 205 and Expedition 301 dealt with this problem by including bridge plugs or packer seals near the bottom of the main casing strings.

• With each new level of resolution provided by technological improvements to sensor design, memory capacity, and power efficiency, new signals have been observed. The earliest instruments provided only 12 bit digital temperature resolution. Applying this to the large dynamic range necessary at their ridge axis sites was adequate for the primary task of characterizing the crustal thermal structure, but many questions remained unexplored. Subsequent advances to a 24 bit thermistor ADC have overcome this problem and have allowed things like tidally modulated formation-fluid flow and bottom water thermal stability to be observed and quantified. Advances in the resolution of pressure have also been realized, in the way of higher resolution, greater absolute accuracy, and greater sampling rates. Increases in pressure resolution (2 orders of magnitude), memory capacity (3 orders of magnitude), and data transfer rates (1.5 orders of magnitude) now permit observations of seismic surface waves, oceanographic infragravity waves, tsunamis, and other subtle signals, both at the seafloor and within the formation. This capability points clearly to the future need for hydrologic observations provided by CORKs to be integrated with other collocated and simultaneous observations such as crustal deformation, seismic ground motion, and seafloor compliance.

Top of page   |   Previous   |   Next