Search

This paper examines numerical approaches to model operation of gas storage in salt caverns. The emphasis is on taking into account the thermal exchanges between the well, the cavern, and the rock mass, as well as the modelling of the volume losses of the cavern, while keeping the model fast, simple, and operationally usable. The use of this model in the context of monitoring storage operations is illustrated.

An experimental characterisation of a low permeability caprock is first reported. Two different methods are compared to measure the porosity: water imbibition and weighting versus gas (Argon) pycnometry. Water measured porosities appear to systematically underestimate gas measured porosities. A capillary pressure curve is derived from successive equilibrium of samples at imposed relative humidities. The relative gas permeability is thoroughly measured on the same samples. Gas porosity and effective gas permeability measurements have been carried out under different confining pressures. All measured relative gas permeabilities lie on a master curve and present a low critical water saturation beyond which the effective gas permeability nearly vanishes. In an attempt to model this last feature, the salient properties of the microstructure are extracted from electron microscopy images to propose a micro-macro model of the relative gas permeability. The multi-scale modeling is performed within the framework of random media or so-called continuum micromechanics, including localised flow effects within interfaces. The self-consistent homogenisation scheme allows retrieval of a critical water saturation, whose value is governed by the shape of the elementary particles of the material and the localisation of water within the pore space.

Abstract Background Adult T cell leukemia results from the malignant transformation of a CD4+ lymphoid clone carrying an integrated HTLV-1 provirus that has undergone several oncogenic events over a 30-60 year period of persistent clonal expansion. Both CD4+ and CD8+ lymphocytes are infected in vivo; their expansion relies on CD4+ cell cycling and on the prevention of CD8+ cell death. Cloned infected CD4+ but not CD8+ T cells from patients without malignancy also add up nuclear and mitotic defects typical of genetic instability related to theexpression of the virus-encoded oncogene tax. HTLV-1 expression is cancer-prone in vitro, but in vivo numerous selection forces act to maintain T cell homeostasis and are possibly involved in clonal selection. Results Here we demonstrate that the HTLV-1 associated CD4+ preleukemic phenotype and the specific patterns of CD4+ and CD8+ clonal expansion are in vivo selected processes. By comparing the effects of recent (1 month) experimental infections performed in vitro and those observed in cloned T cells from patients infected for >6-26 years, we found that in chronically HTLV-1 infected individuals, HTLV-1 positive clones are selected for tax expression. In vivo, infected CD4+ cells are positively selected for cell cycling whereas infected CD8+ cells and uninfected CD4+ cells are negatively selected for the same processes. In contrast, the known HTLV-1-dependent prevention of CD8+ T cell death pertains to both in vivo and in vitro infected cells. Conclusions Therefore, virus-cell interactions alone are not sufficient to initiate early leukemogenesis in vivo.