--> --> Unconventional Reservoir Characterization with Upscaled Permeability Using SEM

International Conference & Exhibition

Datapages, Inc.Print this page

Unconventional Reservoir Characterization with Upscaled Permeability Using SEM


Conventionally geoscientists have used optical microscopy for micro-tectonics, mineralogy, paleontology and so forth. The more deep we go, the more knowledge flourishes and every next level of magnification teaches us more. However there is a strict limitation of optical microscopy; the wavelength of light does not allow us to go further than fractions of a micrometer; as a result, we cannot inspect unconventional fine-grained, tight-fractured reservoirs with optical resolutions. On the other hand, Scanning Electron Microscopy (SEM) provides a good resolution in the scale of a few nanometers; this is fairly enough for even shales. From literature we know that shales are organically rich and fractured rocks. Indeed, SEM studies show that organics are distributed heterogeneously within inorganic matrix and fractures exist randomly in shales. Capabilities of SEM comes with a disadvantage; SEM can only represent a few square nanometers of a rock, but we deal with hundreds meters long reservoirs in oil and gas industry. An upscaled permeability value characterizing the whole heterogeneous and anisotropic shale reservoir can be a practical solution to such problem. SEM images are digitized in grid layouts. Each grid cell correspond one sub-element; organic, inorganic or fracture. Porosity and permeability of each sub-element is defined in a Computational Fluid Dynamics (CFD) simulator and the production rate for that digitized model is calculated. At the next stage, two modules of software we coded generate different scenarios imitating the base model. Such variations include total random fractures and fractal based branching fractures. By testing production rates of newly created models we determine a dependable production rate range because we have obtained different and extreme cases that are possible to be observed in different images of the same sample. And at the final stage, all models are represented with an upscaled permeability value, yielding the best-matched production rate curves. The upscaled permeability value can be used to define the whole reservoir from which the samples are taken. As a result, we eliminate the disadvantages of SEM which can only characterize a couple square nanometers of a reservoir. We suggest a production rate range including the most extreme cases and an upscaled permeability value which can be used to represent the whole reservoir. Hence, we can estimate the economic worth of petroleum reservoirs more accurately.