No abstract

Potential impacts of gas hydrate exploitation on slope stability in the Danube deep-sea fan, AbstractMethane production from gas hydrate reservoirs is only economically viable for hydrate reservoirs in permeable sediments. The most suitable known prospect in European waters is the paleo Danube deep-sea fan in the Bulgarian exclusive economic zone in the Black Sea where a gas hydrate reservoir is found 60 m below the seafloor in water depths of about 1500 m. To investigate the hazards associated with gas production-induced slope failures we carried out a slope stability analysis for this area. Screening of the area based on multibeam bathymetry data shows that the area is overall stable with some critical slopes at the inner levees of the paleo channels. Hydrate production using the depressurization method will increase the effective stresses in the reservoir beyond pre-consolidation stress, which results in sediment compaction and seafloor subsidence. The modeling results show that subsidence would locally be in the order of up to 0.4 m, but it remains confined to the immediate vicinity above the production site. Our simulations show that the Factor of Safety against slope failure (1.27) is not affected by the production process, and it is more likely that a landslide is triggered by an earthquake than by production itself. If a landslide were to happen, the mobilized sediments on the most likely failure plane could generate a landslide that would hit the production site with velocities of up to 10 m s -1 . This case study shows that even in the case of production from very shallow gas hydrate reservoirs the threat of naturally occurring slope failures may be greater than that of hydrate production itself and has to be considered carefully in hazard assessments.

This paper compares 5,715 equivalent linear analyses to examine the sensitivity of the results to the shear strain ratio ( SR). The results show that as SR increases, equivalent linear analyses predict larger values of maximum shear strain ( γ max), peak ground displacement, mean period ( T m), and spectral acceleration ( Sa) at long periods, smaller values of peak ground acceleration ( PGA), Arias intensity, significant duration, and Sa at short periods, and similar values of peak ground velocity and Sa at middle periods. SR has a non-negligible effect on predicted values of γ max, T m, and Sa at short and long periods. This study also compares measured values from 80 ground motions and six vertical arrays with the results of equivalent linear analyses conducted using two methods to calculate SR. The results show that using SR = 0.65 gives a 2–10% better fit than using SR = ( M − 1)/10, where M is the earthquake magnitude.

Offshore landslides pose a major threat to subsea facilities, and one of the main triggering factors of offshore landslides is strong earthquakes. The state of the practice models offshore slope stability during strong shaking using one horizontal component of a selected representative ground motion. However, slopes and earthquakes are three-dimensional phenomena. Therefore, to provide much needed guidance in this area, this study performed 27 three-dimensional seismic slope stability analyses in a finite element program to investigate the change in the predicted results when applying one, two, or three ground motion components. The results showed that applying two horizontal components instead of one may increase the predicted total displacements and shear strains on the slope by 25% to 50% and by 10% to 50%, respectively. In addition, it may also increase the total displacements at the seafloor on the shelf above the slope by 20% to 40%. This could have implications when deciding the safe stand-back line for offshore facilities. However, applying three ground motion components has a negligible effect on the results compared with applying only the two horizontal components. The ratio of the response spectra on the shelf, slope, and basin varied by ± 25%, however no clear trend was observed.

This paper proposes a unified model to estimate the in-situ small strain shear modulus of clays, silts, sands, and gravels based on commonly available index properties of soils. We developed a model to predict the laboratory small strain shear modulus (Gmax,lab) using a mixed effects regression of a database that contains 1680 tests on 331 different soils. The proposed model includes the effect of void ratio, effective confining stress and overconsolidation ratio as well as plasticity index, fines content, and coefficient of uniformity. We compiled a second database to estimate the in-situ small strain shear modulus (Gmax,in-situ) from laboratory (Gmax,lab) measurements. This study validated and compared the resulting model with other existing models using a third database of measured Gmax,in-situ values. The residuals of the proposed model had a mean and median closer to zero and the smallest standard deviation of all the models considered. By including a statistical description of the residuals, this work allows uncertainty of the small strain shear modulus to be included in probabilistic studies.