A formation intelligent evaluation solution for geosteering

Geosteering system is the optimal placement of a three-dimensional well trajectory based on the results of real-time downhole geological, geophysical and drilling parameters, aiming to obtain the largest oil drainage area and the best recovery factor. It has become the cutting-edge technology to improve the oil and gas production of a single well and the benefits of oilfield development, confronting geological problems such as low porosity, low permeability, and strong heterogeneous reservoir, combined with drilling engineering problems such as high temperature, high pressure, and strong vibration. Based on sorting out the hardware, software, and team of the geosteering system, this paper divided the geosteering system into three stages: trajectory geosteering based on formation structure, reservoir geosteering based on reservoir lithology, and productivity geosteering oriented to formation composition. Taking the outcrop with higher resolution and more intuitive as an example, the logical relationship in three levels of "structure-lithology-composition" for the geosteering system was expounded. In this paper, a formation intelligent evaluation solution for geosteering with the scheme of "data acquisition, information fusion, situational awareness, and formation intelligent evaluation" was proposed: According to the discipline categories, combined with the data sources, acquisition time and spatial resolution, the data types and characteristics required for formation intelligent evaluation were sorted; The information fusion classification of multi-source heterogeneous data of geosteering was proposed into "Data level -Feature level-Decision level", and the relevant algorithms were sorted out; According to the data model and prior experience, the situation assessment for the current drilling state and the geological environment was carried out, and the situation forecast for the next drilling state and the geological environment was carried out by machine learning algorithms; According to the "structure-lithology-composition" classification, the concepts of "Preliminary Geosteering" and "Fine Geosteering" was put forward, and the technical points of "Fine Geosteering" were also determined. An application of the geosteering intelligent evaluation solution to the actual oilfield geosteering operation was carried out, verifying its reliability and practicability, which can provide a reference for future downhole intelligent closed-loop research.