TEXAS A&M INSTITUTE OF DATA SCIENCE

Digital Twin Lab

Pilot Project – Digital Twin for Geo-Engineering

Digital Twin for Enhancing Geo-Systems (Drilling and Production) in the Context of Clean Energy Extraction

Project Lead: Eduardo Gildin (Department of Petroleum Engineering)

Meeting the net-zero emission paradigm will require a realignment of hydrocarbon production strategies with other forms of energy production. Carbon Capture Storage and Sequestration (CCS) and Compressed Hydrogen (H2) energy storage along with Geothermal energy production will contribute tremendously to achieving a sustainable form of cleaner energy production in the future.

Profiting from all of these sources of energy is only possible if an accurate and timely prediction of the injection-production behavior of fluids in the subsurface can be attained. This involves the development of drilling and subsurface and surface flow simulators that can generate accurate answers in a timely manner (real-time) using commodity-based platforms (e.g., laptops, tablets, etc.). Meeting the net-zero emission paradigm will require a realignment of hydrocarbon production strategies with other forms of energy production. Carbon Capture Storage and Sequestration (CCS) and Compressed Hydrogen (H2) energy storage along with Geothermal energy production will contribute tremendously to achieving a sustainable form of cleaner energy production in the future. Profiting from all of these sources of energy is only possible if an accurate and timely prediction of the injection-production behavior of fluids in the subsurface can be attained. This involves the development of drilling and subsurface and surface flow simulators that can generate accurate answers in a timely manner (real-time) using commodity-based platforms (e.g., laptops, tablets, etc.).

Geothermal wells share similar drilling and completion challenges with standard oil and gas wells, and yet, many of the recent deployments of advanced controls and automation in oil and gas wells have not permeated through the geothermal industry. Part of the problem is that geothermal wells are more individualized in terms of geological and petrophysical properties (even more than unconventional plays) and therefore depend heavily on good reservoir characterizations. Although this seems a big shortcoming in implementing efficient drilling campaigns, it is in fact, a great benefit to the design of automated drilling processes, in which repeatability is paramount for efficiency. Automated control systems workflows, in this case, can allow the real-time identification and changing of operational parameters during drilling operations. 

In phase I of this pilot project, we will devote time to building a DT for the drilling process. In this case,  a central contribution of the proposed task is the development of a multifaceted framework that combines simulation, data assimilation, automated control, and optimization for coupled nonlinear drilling processes into the DT framework of a drilling rig. . We will therefore build on our previous work, especially combining expertise in modeling hydraulic fracturing and drilling processes to equip our controls and trajectory tracking strategies with estimation algorithms to determine (1) the state of the drill-bit and (2) design optimal ROP, RPM, and plasma discharge to create efficient microcracks into the rock formation. We will be connecting a physical drilling rig (miniaturized drilling rig) developed in previous works with a digital platform simulating a drilling operation. 

In phase II of this project, we will expand the framework to include the reservoir and flow information.  A digital twin is expected to realize interaction and integration between physical space and the virtual world due to its feasibility of real-time synchronization, real mapping, and high fidelity. In order to create the virtual model of the physical entity of either a fossil fuel reservoir or a geothermal reservoir, the fluid flow and transport process in realistic reservoir conditions are mathematically described in the digital space, while data is transferred into the twin and information is fed back from the twin. We will use a commercial off-the-shelf reservoir simulator as our physical entity and will build up proxy models to work as a digital replica of the true reservoir. It is worth noting that the surrogate framework developed in the pilot project on Nuclear Security Applications can be used in the context of reservoir modeling (and vice-versa), bringing lots of synergies between both groups.