Drag embedment anchors (DEAs, or anchors) are used as foundations to secure mobile offshore drilling units (MODUs) in soft clay soils on the sea floor. In 2004 and 2005, Hurricanes Ivan, Katrina, and Rita caused the mooring failures of 17 mobile offshore drilling units moored with anchors. Since then, a great deal of research has been conducted regarding anchor performance and reliability. This report provides an overview of anchor research and industry practice to date, and discusses the results of two research experiments to assess anchor performance. One experiment investigated the effect that embedding anchors in the same soil path several times had on bearing force. The second experiment investigated what effect changing the direction of the horizontal load vector relative to the anchor shank had on embedded anchor bearing force. The results of these experiments suggest that remolding clay may have an effect on anchor bearing force capacity, and that repeatable results are obtainable when testing the effect of changing the direction of applied horizontal load. / text

The failed anchoring systems of mobile offshore drilling units from hurricanes occurring in 2004 and 2005 established a need to better understand the ultimate pullout capacity and trajectory of scaled model anchors under typical and out-of-plane loading conditions. The six degrees of freedom of small scale drag embedment anchors were studied in a laboratory testing environment with the intent that reasonable trends in anchor behavior will be found. Investigations within this experimental research program demonstrated the in-plane and out-of-plane loading behavior of conventional and prototype scaled model anchors embedded to predetermined depths in two different test beds of kaolinite clay with undrained shear strength profiles constant and increasing with depth. The anchors were loaded to failure in concentric, normal, concentric, shear, eccentric, normal, eccentric, shear, inclined, and drag embedment loading configurations. This series of pullout and drag embedment tests provided a suite of test results indicating behavioral trends of the varying holding capacities and anchor trajectories. Results were compared with similar research presented in the literature and an analytical model predicting out-of-plane loading behavior of similar anchors. It was concluded that increasing eccentricities from both concentric, normal and concentric, shear loading configurations resulted in decreasing bearing capacity factors, confirming the predicted trend from the analytical model for these loading configurations. Trajectories observed for the concentric, normal, concentric, shear, and eccentric, shear loading configurations showed that the anchors tracked straight out of the soil without deviation, but eccentric, normal loading found the anchor tending to track away from the initial loading location. For inclined loads, both anchors to track to whichever direction the anchor faced upon loading. Drag embedment trajectory was found to vary depending on the anchor, as the conventional anchor dove with an applied load and the prototype anchor rose towards the surface. / text

The trajectory and capacity are key components of the design of drag embedment anchor and drag-in vertically loaded anchors. This experimental testing program quantifies two factors that describe the anchor trajectory and capacity: the equilibrium bearing factor (Ne) and the tangential bearing factor (Ne). These factors can aid in the development of a numerical model of anchor behavior. A magnetometer device is used to track the orientation and location of the anchor during drag embedment. The results of the experimental testing program were compared with the results from a predictive model. The experimental program consisted of drag embedment tests with various testing conditions including different anchor line diameters and different initial pitch orientations. The results with the different anchor lines indicated that thinner anchor lines cause the anchor to dive deeper in the soil. The different initial pitch results indicate that regardless of the initial pitch of the anchor, the anchor rotates to a unique pitch trajectory within 2 fluke lengths. / text