Using a thin sand layer to ease spudcan extraction in clay

Abstract

Difficulties are encountered, in soft clay and silt sediments, during extraction of 25 the spudcan foundations of 'mobile' jack-up rigs. This paper reports results from centrifuge 26 model tests undertaken to provide a simple, effective and practical means for easing spudcan 27 extraction in soft normally consolidated and lightly overconsolidated clay. The tests were 28 carried out on model spudcans of two different geometries typically used in the field. The 29 spudcans were extracted from penetrations of 2~3 diameters and after an operation period of 30 ~19 months. The measured pore pressure at the bottom face of the spudcan during extraction 31 process provided insight into the development of suction at the interface between the spudcan 32 base and the underlying soil. 33The tests were carried out with and without a sand layer placed locally in-between the 34 spudcan and the mudline at the onset of penetration. In the field, a gravel layer can be placed 35 by means of a flexible fall pipe system. The presence of a sand layer (with a thickness less 36 than the spudcan tip length) beneath the penetrated spudcan assisted in easing leg extraction 37 difficulties allowing the base suction to be reduced by 73~83% and hence the breakout force 38 by 31~32%. For a given penetration depth, the installation resistance was also increased by 39 24~26%. This yielded a 15~17% reduction of spudcan penetration depth under a given 40 preload and hence the breakout force as much as 50%. Guidelines and cautions for using the 41 proposed measure are provided. Additional advantages of the new recommended method in 42 enhancing the efficiency of the conventional measures of lessening spudcan extraction 43 difficulties are discussed. 44

Extraction response of skirted foundations and a spudcan on sand-over-clay deposits

“…It is seen from Figure 4 Typical preload range on 12-m diameter spudcans were 35~70 MN. Although it was not given, the corresponding breakout force can be calculated as 0.7  35~70 = 24.5~49 MN following the experimental results presented by Hossain et al (2015). This range is also shown in Figure 4 and…”

“…t r,e = t r,p = t). The separation of the spudcan base from the underlying sand plug is also confirmed by a snapshot from a halfspudcan test (from Hossain et al, 2015). Note, this is also evidenced through the dissection of the specimen in Box 2.…”

“…A load cell fitted between the loading arm of the actuator and the foundation shaft measured load-penetration responses in this displacement controlled system. Hossain et al (2015) and the evidence is shown in Figure 5. Figure 5 displays the sand plug left in the clay layer by dissecting the layered soil specimen in Box 1 after the extraction of the foundations.…”

“…If there is a naturally deposited sand layer above that clay layer or a sand layer is spread artificially (Case II), the spudcan penetration depth (and hence required jack-up leg length) and breakout force could be reduced Hossain & Dong, 2014). Hossain et al (2015) showed that an addition of 2 m sand layer has yielded a 15~17% reduction of spudcan penetration depth under a given preload and hence the breakout force as much as 50%. However, depending on the relative thickness (t/D) and density (I D ) of the sand layer and the undrained shear strength of the underlying clay layer, the spudcan may experience a punch-through incident, which can be assessed using the methods recommended by the recently finalised version of ISO guidelines 19905-1 (ISO, 2012) or the methods proposed by Lee et al (2013b) and Hu et al (2015a).…”