Design of a free-fall penetrometer for geotechnical characterisation of saturated sediments and its geological application

Abstract

Cone Penetration Testing (CPT) is a versatile, time efficient method to characterise sediment strength and pore pressure in offshore settings and on land. The majority of the penetrometers rely on heavy trucks or rigs to provide the necessary force to push the CPT probe into the ground. This laborious process usually deforms or otherwise affects the uppermost deposits, whose physical properties are in turn vital to understand processes related to scour, burial of mines, cable or pipeline laying, silting of water ways and harbours, or sediment transport and remobilisation. Owing to the shortcomings of heavy seagoing CPT gear, this thesis aimed to develop and use a cable-led marine penetrometer lance which profiles the uppermost sediments in a less destructive manner. The study summarises the development and deployment of a marine cone penetrometer system. It consists of an instrument for shallow-water application (200 m) and, based on the experience of the first, a second version operable down to 4000 m water depth. Design and construction of the instruments occurred at the Research Centre Ocean Margins, Bremen University (RCOM) in close and productive collaboration between Achim Kopf, Matthias Lange and myself during the first year. After an initial phase of researching for sensors and components, I contributed to the design. After construction, a total of 338 CPT experiments (both with the shallow-water [SW] and deep-water [DW] device) were carried out to date. From the wealth of deployments, 300 of them were performed by me while 204 raw data sets were processed by me over the course of this thesis (mostly 2nd and 3rd year). The CPT deployments were initially dedicated largely to instrument testing, and later focused on geological application. From 38 selected data sets, the strain-rate effects of the dynamic tests were assessed and, based on earlier empirical solutions from the CPT literature, corrected for. Overwhelmingly, the results with the RCOM lance agree well with those from the pushed tests, and further help to accentuate them. Within the spectrum of velocities tested , it is found that the faster the rate of initial penetration is chosen, the larger the discrepancy between the deviations caused by layering and variations in physical properties of the sediments. This observation may be vital when carrying out CPT experiments in geomaterials where lithological variability is small, because it helps identifying features otherwise undetectable by pushed profiling at the standard rate of 2 cm/s. Another five manuscripts summarise the geological application of the CPT devices. The tests were performed in geological environments as diverse as the Baltic Sea, Lake Lucerne, an active mud volcano in Azerbaijan, and the Cretan Sea in the Eastern Mediterranean. Regardless of the regional scenario, some overarching consistent results were obtained during the CPT deployments. As an outstanding finding, three types of characteristic pore pressure signals are recorded by CPT lance when deployed in "free-fall" mode on a cable. In granular, normally consolidated material, a pore pressure spike upon impact is usually followed by an exponential decay back to ambient values (if sufficient time is allowed for dissipation). Alternatively, a second pattern often met is a negative (i.e. sub-hydrostatic) pressure spike followed by an increase in pressure to ambient pore pressure values. The sub-hydrostatic signal is caused by displacement of pore fluid by the profiling CPT instrument, which results in flow away from the probe. This second pattern is restricted to coarse-grained deposits with high permeability. The third characteristic signal generally shows supra-hydrostatic pressures upon impact and during profiling, but then climbs to even higher pressures with time. Graphs like this are found in sediments of variable grain size distribution and are related to fluid overpressures. Interestingly, the third type is observed in clays as well as silt- or sand-bearing deposits, and appears no matter what the cause of the excess pore pressure is. In the various field studies, very similar pore pressure curves are seen although the reason for the overpressures were glacial loading (and potentially EQ tremor; Lake Lucerne), presence of microbial gas (Baltic Sea), hydrocarbon formation in folded and faulted shales of the Maykopian Formation (Azerbaijan, Greater Caucasus), or neo-tectonic movement and landsliding (Cretan Sea). In summary, this study has shown that velocity-controlled "free-fall" CPT lances are an efficient, user-friendly means to characterise geotechnically shallow sub-bottom deposits. They obtain reproducible results that can be linked with standard pushed tests, but have the added advantage of producing more pronounced excursions in cone resistance and sleeve friction as well as characteristic pore pressure responses indicative of geological conditions.