This website uses cookies to improve your experience. If you continue without changing your settings, you consent to our use of cookies in accordance with our cookie policy. You can disable cookies at any time.

×
On Tuesday, 28 May 2024, from 1:00 am CDT to 1:00 pm CDT, the SEG Library will undergo a major site upgrade. During this time, users will be able to access their accounts, but certain features may be unavailable. We apologize for the inconvenience.

Can we reliably estimate brittleness for thin shale reservoirs? A case study from the lower Paleozoic shales in northern Poland.

Authors:

Abstract

Here, we apply a workflow that allows us to differentiate brittle and ductile rocks of the Lower Paleozoic shale play in Northern Poland (Baltic Basin) using surface seismic data. We start from the brittleness evaluation using well logs by applying both: 1) rock-elastic and 2) mineral content-based methods. As the well log based brittleness gives us information only near the borehole, we estimate brittleness of the reservoir in three dimensions at seismic scale by integrating well logs and set of rock-elastic property volumes that are calculated based on the simultaneous AVO inversion results. For elastic seismic brittleness estimation we apply the combination of Poisson’s ratio and Young’s modulus, whereas for mineral content-based brittleness we use a proximal support vector machine algorithm (PSVM). Because of the small thickness of the target intervals (< 25 m) as well as low resolution of the input data, we are not able to obtain satisfactory results in case of the seismic elastic-based brittleness calculations. Therefore, we calculate the seismic mineral-based brittleness by applying an alternative approach – supervised learning classification algorithm – to achieve better results. The main problem with this method is related to the limited number of available wells to train the algorithm and validate it, as well as the classification itself. Despite this fact, the mineral brittleness predicted from seismic using PSVM, shows more details of the target formations as compared to the mechanical one. We have a strong belief that in the case of thin shale reservoir as well as low resolution of the input seismic data, the PSVM is a preferred solution to predict volumetric brittleness.

Presentation Date: Tuesday, September 26, 2017

Start Time: 9:45 AM

Location: Exhibit Hall C/D

Presentation Type: POSTER