Chapter – Conclusions and Further Research Chapter Conclusions 7.1 Introduction A footprint is characterised by the spudcan imprint geometry left on the seabed and the soil condition underneath it. The spudcan footprint characteristics and its influence on spudcan-footprint interaction are studied and presented in Chapter 4. The soil condition is found to change with time as the soil re-consolidation takes place. This is especially crucial for spudcanfootprint interaction in soft clay where the soil shear strength variation dominates the interaction. The changes in soil properties are also investigated in this study and presented in Chapter 5. As spudcan-footprint interaction is essential a soil-structure interaction issue, the interaction is not only affected by soil condition but also the structural configuration. The results of the study on the effect of structural configuration are presented in Chapter 6. The main achievements of this study are: 1) the identification of the pre-dominant factor of footprint characteristics on spudcan-footprint interaction in clays of varying strength profiles, 2) the importance of soil reconsolidation with time and its implication on spudcan-footprint interaction, and 3) the development of charts for predicting the maximum horizontal force and moment induced on spudcan and the corresponding depth of occurrence. These key findings have been achieved through an extensive programme of 265 Chapter – Conclusions and Further Research centrifuge modelling tests. The main findings of this study are elaborated in the next section. 7.2 Main findings 7.2.1 Characteristics of spudcan footprint and its influence on spudcan-footprint interaction Spudcan footprints have been commonly defined as seabed depressions left by the previous jack-up rig (SNAME, 2002a, b) with no sufficient emphasis on the importance of soil condition underneath the depression. In this study, the spudcan footprint is specifically characterized by not only the seabed depression profile but also the soil condition underneath the depression. Centrifuge model tests were conducted to investigate these footprint characteristics on clays of varying strengths ranging from firm to soft. Owing to shallow penetration in firm clay, the footprint formed is likely a depression of one spudcan diameter wide with an almost vertical cylindrical circumference and a relatively less disturbed base. Owing to deep penetration in soft clay, the footprint formed is likely a bowl-shaped depression with a width of spudcan diameters and soil heave along the footprint periphery. In general, the soil experiences substantial reduction in su up to a distance of 0.8D to 1.0D from the footprint centre, with the reduction decreases with increasing radial distance. With negligible soil reconsolidation (in the shortterm), the su contours basically follow the physical profile of the depression explaining why the tendency of the spudcan to slide towards the footprint centre. The physical profile of the depression is recognized to be the dominant factor for spudcan-footprint interaction in firmer soils, whereas the soil 266 Chapter – Conclusions and Further Research strength variation is found to be more significant in softer soils. This finding has a significant implication in devising an effective mitigation measure to overcome the problem of the spudcan sliding into the footprint. It is obvious that infilling footprint crater with some imported materials in cases where spudcan-footprint interaction cannot be avoided as recommended by the guideline SNAME (2002a, b) is not feasible to mitigate footprint problem in soft clay. In term of soil failure mechanism during spudcan-footprint interaction, a half-spudcan test was performed. The result reveals that soil bearing failure is likely to occur for penetration above the crater depth, whereas soil sliding failure is likely to dominate the soil failure mechanism when the spudcan penetrates further below the crater till the penetration depth of the previous spudcan. The mechanism identify may help future studies to investigate the effectiveness of different mitigation methods such as stomping, skirted spudcan, infilling etc. 7.2.2 Effect of time on spudcan-footprint interaction As found in the earlier section, the soil strength variation in a footprint is dominant for spudcan-footprint interaction in softer clays. The degree of soil strength variation changes with time as excess pore pressures generated by the previous spudcan installation dissipate (Gan et al, 2007 and Leung et al., 2007). The soil undergoes re-consolidation during the operational period and after the extraction of spudcan. In addition to time, the change in soil strength of a footprint is also affected by the soil initial stress state, the stresses induced during the spudcan penetration and extraction. In NC clay, the initial soil condition of a footprint consists of a deep region of heavily remoulded soil 267 Chapter – Conclusions and Further Research with a width of 1.5D (or 0.75D from the footprint centre), where D is the spudcan diameter. The soil gains strength of up to twice the undisturbed soil when the excess pore pressure generated has fully dissipated. The soil further away from 1.5D from the footprint center is likely to remain intact with no significant disturbance from the spudcan activity and no significant strength gain in the long-term. A ‘localized crust’ beneath the spudcan base is formed as the soil undergoes consolidation during the operational period. The strength of this crust increases with the duration of jack-up operational period. This crust induces detrimental moment on the spudcan that is installed partially overlapped with it due to considerable variation in bearing resistance at this level. The findings suggest that a lower preload pressure rig may be a better choice for this type of soil condition as to ensure the rig penetration can be terminated at a sufficient depth above the crust layer. The initial soil condition of a footprint formed in OC clay is found to be similar to that observed in NC clay. In the short-term (when the soil reconsolidation is negligible in both operational period and elapsed time after a footprint is formed), the extent of soil disturbance corresponds well to the soil failure mechanisms during the spudcan penetration and extraction. However, in the long-term (when considerable length of time is involved), the soil condition in OC is clearly different from that observed in NC clay. Relatively less shear induced pore pressure is generated in a footprint formation in OC clay compared to that in NC clay. Hence, though the soil gains some strength in the long-term, the shear strength is generally lower than that of the undisturbed soil. The spudcan experiences a relatively high H and M at depth above de in the short-term and below de in the long-term. This finding suggests 268 Chapter – Conclusions and Further Research the adverse effect of a footprint varies with time as the soil condition changes. Carefully selecting a suitable rig and positioning it at a safe distance from footprint interference is recommended. Parametric studies on the effect of leg flexural rigidity, spudcan diameter, preload pressure and offset distance on spudcan-footprint interaction were performed and reported in Chapter 6. The findings can be used as a general guideline for selecting a suitable rig with respect to the field condition. 7.2.3 Parametric studies and dimensional analysis for spudcanfootprint interaction A generalised short-term soil condition of a footprint in soft to firm clay, which was developed based on test results (3 tests were done in NUS and tests were done in UWA), is shown in Fig. 6.1. Details of the generalised footprint condition can be found in Section 6.1. Parametric studies were performed to investigate the influence of various parameters such as leg flexural rigidity, spudcan diameter, preload pressure, and offset distance, on spudcan-footprint interaction. No distinguished difference in induced H profile for the leg stiffness tested in a range of 7.82×1010 to 1.22×1012 Nm2. The restrained connection used in the model set-up may attribute to this result as the effect of horizontal displacement on the soil reaction is discounted. On the other hand, higher H and M were generated for larger spudcan diameter demonstrating the induced forces are a function of spudcan diameter. In term of preload pressure, higher forces were induced for case with higher preload pressure of the first installation as deeper remoulded soil region was created. Dimensional analysis was performed based on some earlier findings and empirical relationships to estimate the potential maximum forces (Hmax 269 Chapter – Conclusions and Further Research and Mmax) induced on the spudcan during the interaction at an offset distance of 0.5D were proposed. Charts showing these relationships in dimensionless forms are presented in Figs. 6.10 – 6.11. Similarly, the depth of occurrence of these forces was also studied. It is found that the Hmax and the Mmax are likely to occur at a similar depth that is about 0.2 spudcan diameter above de. In term of offset distance, relatively high H and M were induced on the spudcan for the re-installation within 0.5 to 1D offset from the footprint centre. The findings suggest that these offset distances should be avoided to reduce the risk of spducan sliding and structural over-stressing. In term of size ratio Df/Ds (a ratio of footprint diameter to spudcan diameter), higher Hmax/suDs2 is observed for cases with higher Df/Ds which the previous rig has larger spudcans than the future rig. The study has identified some critical parameters such as spudcan size, preload pressure, offset distance and spudcan size ratio, which affect the spudcan-footprint interaction and need to be considered for future rig deployment. The established relationships provide some guidelines for a better rig selection and rig positioning. 7.2.4 Limitations All experimental results on spudcan-footprint interaction presented in this study were obtained from the centrifuge model tests of a single leg with a fixed connection between the top of the loading leg and the centrifuge actuator. This set-up is closer to a scenario where only one leg of a jack-up rig interacts with an existing footprint and the other legs are pre-installed on the intact ground. Stewart & Finnie (2001) pointed out the load imposed on the spudcan during spudcan-footprint interaction is a function of horizontal displacement and vice-versa. With the fixed connection, the lateral 270 Chapter – Conclusions and Further Research displacement and rotational movements were restrained and the forces measured are deemed to be in a high side. The empirical relationships proposed in this study were established based on limited experimental data with some prescribed test conditions; the accuracy proposed empirical relationships may be discounted. However, the footprint problem is complex in nature and generally lack of a proper site investigation for future rig installation (due to high cost and time constraint), the proposed empirical equations provide a channel for estimating the induced forces with a minimum input parameters required (generally, the parameters can be obtained from the previous rig installation record). It is therefore believed that the product of the dimensional analysis can be a useful tool to the jack-up industry for predicting the spudcan foundation response involving spudcan-footprint interaction. 7.3 Recommendations for future research 7.3.1 Soil behaviour and V-H-M response Based on the findings on the combined effects of remoulding and reconsolidation of the soil condition underneath and surrounding a spudcan footprint, a more refined analysis on spudcan-footprint interaction is recommended in order to develop a quantitative framework to better estimate i) soil strength profiles after penetration and reconsolidation and therefore ii) V-H-M responses during spudcan re-penetration. 7.3.2 Footprint mitigation methods Based on the findings established in the present study, the mentioned mitigation methods (e.g. infilling, stomping, Swiss-cheesing and skirted 271 Chapter – Conclusions and Further Research spudcan) may not be suitable for all footprint conditions. Hence, each mitigation method should be studied based on its functionality and feasibility in some specific footprint conditions. Skirted spudcan A skirted spudcan consists of a shallow foundation with a thin skirt at its circumference. Skirted foundation has been extensively studied on its ability of improving foundation stiffness for resisting the horizontal environmental loads (e.g. Svano and Tjelta, 1996; Bransby and Houlsby, 1999; Cassidy et al., 2004) and penetration behaviour in layered soil (Teh et al., 2008; Swee, 2009). In footprint problem, skirts may provide additional lateral resistance against sliding towards footprint. As a rigid connection was adopted which might discount the mobilisation of the soil lateral resistance, Teh et al. (2006) found that the maximum moment that developed on both spudcan and skirted foundation was comparable, and the horizontal loads on these two different foundations acted in the opposite direction. In view of this, a more flexible connection system is recommended for investigating sliding resistance of the skirted spudcan. Levelling out the crater by mean of infilling with suitable material Infilling of footprints formed in a coarse grained seabed with granular materials has been performed in the practice and this should not pose problems as the material characteristics are similar (Jardine et al., 2001). The suitability of this method in fine grained seabed is still in question. For a footprint problem which the depression profile dominates the interaction, the effectiveness of this method relies on the magnitude of differential bearing resistance between the site material and the infilled material. The infill 272 Chapter – Conclusions and Further Research properties may depend on the material used and the placement procedures such as the method of infill and water depth. A further study on this method is recommended despite the comment made by Jardine (2009) based on the numerical analyses results that the crater infilling solutions to footprint problems may not be advisable when considering clay foundations. 7.3.3 Partially restrained leg-hull connection Most of the studies on spudcan-footprint interaction adopted a single leg with fully restrained connection. It has been illustrated by Stewart and Finnie (2001) that the horizontal load-displacement response would be dependent on the structural feasibility. In reality, a jackup leg-hull connection is neither fully rigid nor fully flexible. The basic function of leg-hull connection is to allow forces to transit between the legs and the hull. Each connection commonly consists of a pair of upper and lower guides and a jacking system and/or fixation system (SNAME, 2002b). This kind of connection allows some extent of movements. Hence, experimentally investigation on spudcan-footprint interaction using partially restrained connection between the model leg and the centrifuge actuator is recommended. 7.3.4 Numerical simulation of spudcan-footprint interaction One of the main challenges on the numerical simulation of spudcan-footprint interaction is to properly model the footprint condition including the depression profile and the soil shear strength profile underneath it. In this study, the characteristics of a footprint on clay of varying strength and the behaviour of soil condition changes with time have been established. These results could be useful for the numerical simulation of a more realistic 273 Chapter – Conclusions and Further Research footprint condition. Eulerian finite element method as used by Tho et al. (2009) for analyses of spudcan penetration in clay could be used for simulating spudcan-footprint interaction. Furthermore, the numerical simulation can be extended to study the system behaviour of a jack-up rig interacting with footprints. 274 . investigation on spudcan-footprint interaction using partially restrained connection between the model leg and the centrifuge actuator is recommended. 7.3.4 Numerical simulation of spudcan-footprint interaction One. characteristics on spudcan-footprint interaction in clays of varying strength profiles, 2) the importance of soil reconsolidation with time and its implication on spudcan-footprint interaction, and 3). investigated in this study and presented in Chapter 5. As spudcan-footprint interaction is essential a soil-structure interaction issue, the interaction is not only affected by soil condition but also