Precruise planning for the ACORK in Hole 808I emphasized packer and screen placement for long-term hydrogeologic monitoring within two zones: the upper oceanic crust and the décollement zone. The original planned ACORK configuration included five packers and six screens configured to focus on these two zones as follows: (1) a screen at oceanic basement with a packer immediately above and (2) a symmetric four-packer, five-screen array centered on the décollement zone. This was an ambitious plan, a plan that would have involved deployment of the third deepest DSDP or ODP casing ever, focused on the top-priority targets for studying the relationship of fluid flow and deformation at the toe of the Nankai accretionary prism.
During the leg, these plans were modified in accordance with operational realities for two reasons. First, as is fully described in "Operations" during the LWD operations extremely poor drilling conditions were encountered immediately below the décollement zone. This required the ACORK configuration to be modified to eliminate the basement objective and penetration below the décollement zone. Second, the installation of the 728-m-long ACORK in Hole 1173B encountered increasingly difficult drill-in conditions in the deepest 100 m. A number of hypotheses were considered to explain this behavior and the most likely seemed to be the drag encountered at the upper limit of the lower Shikoku Basin deposits by the array of packers ultimately emplaced across the stratigraphic equivalent of the décollement zone. As the packers are the largest diameter components in the ACORK, we reasoned that limiting the number of packers would be required in the even deeper installation at Site 808.
Given these constraints, a number of options were considered for the actual ACORK configuration in Hole 808I, ranging from a short, four-packer, five-screen installation focused only on the shallow frontal thrust zone to a multipacker, multiscreen configuration at and immediately above the décollement zone. In the end, we decided to utilize six screens and only two packers, one to be set in the 20-in casing and one immediately above the décollement zone. We reasoned that the décollement zone was the most likely zone to be overpressured and therefore the most important zone to isolate with a single packer in the formation. From an operational standpoint, placing the single open-hole packer as deep in the string as possible was thought to minimize the risk during installation, as it would follow as closely as possible behind the underreamer used to run the entire string into place. We also reasoned that the formation would eventually collapse around the other screens not individually isolated by packers.
Hence, the two packers and six screens were assembled in a 964-m-long ACORK string, to emphasize long-term observations of pressures in three principal zones, as follows:
As described in more detail in "Operations" drilling conditions during installation of the ACORK steadily worsened starting at ~200 m above the intended depth. Despite a heroic effort by the rig crew and application of every available technique, progress stopped 37 m short of the intended installation depth. This left the screen sections offset above the intended zones (Fig. F1; Table T3), not an ideal installation but still viable in terms of the scientific objectives. In addition, this left the ACORK head 42 m above seafloor, again not an ideal situation because engineering calculations indicated that the exposed ACORK casing string was probably not strong enough to support its own weight. Indeed, when the drill string was pulled out of the ACORK, the VIT camera feed showed the ACORK slowly tipping over within seconds (Fig. F31).
But the Nankai dragon was feeling generous or perhaps was sleeping! Amazingly, the exposed ACORK components tipped gently and in the best possible direction. Careful inspection with the VIT camera (Fig. F32) showed the casing was not broken but bent continuously over the rim of the reentry cone and a large pile of cuttings. The orientation was such that the critical hydraulic umbilical was undamaged and the components shown in Figure F8A were exposed. This left the data logger and, most important, its underwater-mateable connector in an excellent orientation for future data download by ROV or submersible. Note that the collapse of the above-seafloor section obviously precluded installation of a bridge plug as originally planned, but the bottom of the hole had been filled with heavy mud in anticipation of this outcome (see "Operations").