From hydrate risk management to full-unit optimization, Hariel Mendes, Business Development Specialist at ESSS O&G, details how a Digital Twin ecosystem is reshaping oil & gas production through real-time optimization and advanced simulations.
Optimizing production in oil and gas fields, especially in complex environments like Brazilian offshore fields, demands deep data integration and predictive models. ESSS, in strategic partnership with Petrobras, has developed a Digital Twin solution that extends beyond integrity monitoring, directly focusing on production enhancement through data orchestration and advanced simulations.
The challenge of real-time optimization
Maintaining an updated database of simulation models and providing reliable estimates of production-related parameters is a significant challenge for engineers and managers. Overcoming these hurdles allows for production optimization, loss avoidance, and management of Flow Assurance and Artificial Lift risks.
The central objective of this ESSS Digital Twin is to orchestrate data, optimize the resolution of production problems, and drive process automation, providing operational intelligence. To achieve this, it was designed to assist in critical production tasks, such as generating and calibrating simulation models, optimizing total unit production, evaluating hydrate formation risks, and other processes.
The architecture of the Digital Twin for production enhancement and data orchestration
This Digital Twin ecosystem is composed of various digital applications that manage production-related data from a large number of wells and their respective Floating Production Units (FPUs). Its data orchestration capability is central to its operation:
- Comprehensive Automated Data Collection and Contextualization: The Digital Twin solution ensures automated access to all relevant data. Specific applications aggregate data about wells (trajectory, tubing and casing strings, well characteristics) and subsea pipelines (geometry, diameters, thermal parameters), accessible via Application Programming Interface (API). A Plant Information Management System (PIMS) provides real-time sensor data time series (pressure and temperature from downhole – PDG, wellhead – TPT, and upstream of the choke valve) and periodic well-testing data (Gas Oil Ratio – GOR, Water Cut – WC, flow rate). Another application contextualizes this data, mapping variables to wells and wells to production units.
- Advanced Algorithms for Optimization: The ecosystem hosts a set of algorithms, accessible via API, that perform specific tasks:
- Model Calibration: Calibrate well simulation models to match parameters acquired during well tests.
- Response Surface Curve Generation: Create curves where flow rate is a function of upstream choke pressure, gas-lift flow rate, and Electric Submersible Pump (ESP) rotational speed.
- Total Unit Production Optimization: Optimize the total oil flow rate of a Production Unit by choosing optimal parameters without violating constraints such as maximum Production Unit’s processing capacity, total available gas-lift flow rate, maximum gas-lift flow rate per well, among others. The optimization algorithm is designed to optimize total platform production, whether production comes from satellite wells, manifolds, or both.
- Transient Simulator Model Execution: Allows for the execution of transient simulations.
- Simulator-Agnostic Models: The Digital Twin’s central application allows users to input data to create “agnostic models”. These models contain all the information required to generate a simulation model for any of the multiphase flow simulators used by the Digital Twin. The conversion from an agnostic model to a specific simulation model is done “on the fly” whenever needed to run a process that requires simulation.
- Process Orchestration: Process orchestration is performed by the Digital Twin’s central application, which triggers processes when a predefined condition is met. Triggers are configured by the user, and could be, for example, a change in a monitored variable read from PIMS or the completion of an upstream process in the workflow. Input and output validation for each process could be set to be automatic or manual, allowing for automation of the entire workflow or leaving parts dependent on human intervention. For example, optimizing Floating Production Unit (FPU) production when new well test data becomes available triggers the calibration of that well’s simulation model, the generation of a response surface curve, and its addition to the set of most recent curves from other wells, which then allows running a new optimization process for the FPU.
Implementing the ESSS digital twin isn’t just a technological upgrade, it’s a transformation in how assets are managed. The tangible benefits are clear:
Achieved results and benefits
The Digital Twin ecosystem for Data Orchestration and Production Optimization at Petrobras has proven to be a valuable tool for engineers in their daily activities. It facilitates tracking simulation models and results, providing important information on how to optimize oil production and manage hydrate risks. With the available information, managers and engineers might act more consciously and make better decisions.
Specifically, the system enables:
- Simulation Model Generation: The ecosystem generates the necessary simulation files for multiphase flow simulators.
- Total Unit Production Optimization: The system optimizes the total oil flow rate of an FPU by adjusting parameters such as gas-lift flow rate, wellhead pressure, and ESP frequency for each well.
- Hydrate Risk Evaluation: It is possible to launch transient simulations to generate data on thermal-hydraulic behavior after a well shut-down, aiding in hydrate risk assessment.
Challenges and next steps
In the continuous enhancement of the Digital Twin, significant challenges have been overcome, and others are actively being addressed. For instance, the trustworthiness of sensor data is managed through the use of heuristics, ensuring data quality.
Future developments include improvements to the optimization algorithm to incorporate design constraints, the creation of a real-time digital representation of the well where valve positions and flow parameters can be graphically visualized, and the inclusion of machine learning algorithms for detecting anomalous behaviors (such as pipeline ruptures or sensor failures).
Interested in exploring how ESSS’s Digital Production Optimization can enhance decision-making and asset performance? Talk to our specialists for a detailed technical discussion.
Reference: Resende, R. P., Araujo, J. C., Ribeiro, J. L., Melo, A., Perovano, T. G., Del Puppo, A., Ciambelli, J., Miyatake, L. K., Postal, A., Mendes, R., Curto, H., & Dalla, L. (2025). Digital Twin Innovations in Artificial Lift and Flow Assurance: Overcoming Brazilian Offshore Challenges. Offshore Technology Conference (OTC-35657-MS), Houston, TX, USA, 5-8 May. DOI: 10.4043/35657-MS.
