Sparse Subsurface Sensor Signal Estimation for Formation Damage Assessment via a Smart Orthogonal Matching Pursuit

Carbonate reservoirs exhibit water front movement through microfractures, corridors, and related fracture channels (larger than 5 mm in size) as well as the matrix structure, exhibiting generally complex flow patterns. It is crucial to identify the water front motions and fracture channels inside the flow corridors in order to maximize sweep effectiveness and boost hydrocarbon recovery. Here, we provide a new AI-driven orthogonal matching pursuit (OMP) technique for detecting water front movement in carbonate reservoirs determining possible formation damages that impact the flow within the formation. In order to identify and extract possible fracture channels, the technique first applies a combined artificial intelligence (AI) AI-OMP methodology. After that, a deep learning strategy is used to estimate the water saturation patterns in the fracture channels and assess the resulting formation damage. To identify the fracture channels affecting each particular sensor, the OMP uses the sparse fracture to sensor correlation. The deep learning approach then makes use of the fracture channel estimations to evaluate the patterns of the water front. On a synthetic fracture carbonate reservoir box model with a complicated fracture system, we tested the AI-OMP framework. In order to improve reservoir monitoring, essential reservoir characteristics (such as temperature, pressure, pH, and other chemical parameters) will be sensed using Fracture Robots (FracBots, around 5mm in size). A wireless micro-sensor network is used in this technology to map and track fracture channels in both conventional and unconventional reservoirs. Since magnetic induction (MI)-based communication demonstrates extremely stable and continuous channel conditions with a suitable communication range inside an oil reservoir environment, the system enables wireless network connectivity via MI-based communication. The base station layer and the layer for FracBot nodes make up the two levels of the network's system architecture. To capture data that is impacted by variations in water saturation, many subsurface FracBot sensors are injected in the formation fracture channels. To enhance sensor measurement data quality and better track penetrating water fronts, the sensor placement in the reservoir formation can be modified. They spread out in the fracture channels and move with the injected fluids as they begin to sense the conditions of the environment including formation damage that impact the waterfront movements. They then communicate the data, including their location coordinates, among one another before sending it in a multi-hop fashion to the base station installed inside the wellbore. An aboveground gateway and a large antenna make up the base station layer. To be processed further, the FracBots network data is sent to the control center via an aboveground gateway. In properly identifying the fracture channels and the saturation pattern in the subsurface reservoir, the findings showed high estimation performance of the saturation and the derived formation damage. The findings show that the framework operates well, particularly for fracture channels that are quite shallow (approximately 20 m from the wellbore) and have large variations in saturation levels. As a result, in-situ reservoir sensing may be used to follow fluid fronts and identify fracture channels in a reservoir as well as the arising formation damage. A key element in the data processing and interpretation of the subsurface reservoir monitoring system of fracture channel flow in carbonate reservoirs is presented by the innovative framework. The findings show that in-situ reservoir sensors are capable of providing precise tracking of water-fronts and fracture channels in order to maximize recovery.