Greg Myers, of GE Oil & Gas, explains how the digital transformation of offshore assets enables a competitive advantage by reducing excessive inspection and maintenance costs and increasing uptime with real-time operational data.
 |
|
Fig. 1. System Overview for the marine riser digital twin. Images from GE Oil & Gas. |
To compete in today’s climate of energy price instability, operators and contractors can benefit from data-driven solutions to increase their operations visibility, optimize lifecycle management and reduce costs of maintaining offshore equipment. Drawing critical insights from operational data is a crucial first step.
The cost drivers of offshore drilling can include: (1) unplanned downtime associated with drilling in ultra deepwater or harsh metocean conditions, such as strong loop currents, and (2) resources spent to transport and inspect riser joints onshore. These factors pose challenges on efficiently deploying and maintaining marine risers over their 20+ year lifespan, while meeting safety and regulatory needs. These factors can also cause riser curvature, hang-off deflections and high tensions at the blowout preventer and wellhead, which impart fatigue loading damage on the equipment.
Value from digital transformation
GE has addressed these challenges by creating a digital model of a physical asset, called a digital twin, which can then be optimized to drive enhanced business outcomes. In this instance, both a physical and a virtual model of a marine driller riser was created with the goal of reducing downtime or optimized inspection schedules of these critical assets, which can extend miles in length below the ocean’s surface. The digital twin allows GE Oil & Gas to provide the driller with data-based diagnostics and insights into what is happening to the asset during a regular day or even during an extreme event, and then pick the most optimal way to run that asset. GE has already deployed thousands of digital twins in other applications such as aircraft engines, wind farms and power plants using Predix, the cloud-based operating system it developed for the Industrial Internet.
This marine riser digital twin will allow operators to maintain equipment in a more efficient manner. It can help reduce excessive maintenance and costs of the main tube for drilling contractors with a data-driven approach: a baseline database for drilling riser fatigue damage based on field operational data, in situ riser measurement data and environmental conditions. This “predictivity” allows transition from time-based to performance-based maintenance.
In addition, the system may take the guesswork out of decisions such as when to cease drilling activities due to strong currents or other adverse metocean conditions, or when it is safe to restart operations.
Anything that can provide significant reductions to unplanned downtime is critical. Working with RPSEA [Research Partnership to Secure Energy for America], NETL [National Energy Technology Laboratory] and GE customers, GE’s concept was to give the drilling contractors and operators solutions to increase visibility into the health of their equipment below the surface. The system will help operations and engineering teams respond quickly to potential issues as they occur in real-time.
System approach to the marine riser digital twin
 |
|
Figure 2. Illustration of the clamping system for deployment of the subsea sensing module. |
The digital twin provides near real-time condition monitoring and fatigue estimation of drilling risers (Figure 1). A modular approach was used for designing the subsea platform. The platform consists of an acoustic modem and transducer, rechargeable batteries, tri-axial accelerometers and gyroscopes, and a micro-processor for data acquisition and processing.
Unlike conventional techniques such as strain gauges for direct strain/stress measurement, the system measures the vibrations using accelerometers at select joints along the drilling riser. It transmits the vibration data and other sensor data, such as ocean currents, via acoustic telemetry in near real-time to a topside data acquisition system on the drilling vessel. Topside, advanced machine learning techniques, coupled with a physical asset model of the entire riser string, are used to calculate fatigue life estimates for all riser joints.
The digital twin, or machine learning model, is a virtual model of the riser that is continuously updated against the sensor measurements and metocean conditions. When this data is beyond the original training conditions, a model retraining is automatically triggered. This continuous learning and update of the digital twin model allows us to provide more precise calculation of the riser fatigue life and enable optimizing operations in near real-time. Visualization and alerts are provided by software, which ingests and performs advanced analytics on the field data enhance the operational decision-making for the drillers and operators.
A proof-of-concept of the digital twin was recently deployed in the Gulf of Mexico on a semisubmersible ultra deepwater drilling rig for a nine-week test. Five subsea sensing modules were installed at key joints along the riser string by a remotely operated vehicle (ROV) controlled from the topside, with the lowest module located at a depth of 6200ft on the drilling riser string and the remaining equally spaced from the sea surface, up to 350ft below sea level. The topside data acquisition system with the digital twin software collected processed sensor data every hour and analyzed the riser health and fatigue damage in real-time.
To deploy the subsea sensor module, a two-part clamping system was designed, which is riser agnostic and easily retrofittable onto existing risers. The system consists of two elements, as shown in Figure 2: 1) a clamping element which directly clamped onto an auxiliary line of a riser joint, and a bucket that contains the sensing element and has a rigid attachment to the riser; 2) a sensing element which inserts into the bucket to join the clamping element. This system allows the user to deploy and retrieve the sensor clamp and sensing element either manually or by ROV.
Figure 3 shows the docking of the subsea sensing module inside the sensing element into the bucket of the clamping element by a ROV, and Figure 4 shows the topside acoustic receiver for acoustic data acquisition on the drilling rig platform.
Novel aspects of the marine riser digital twin system from this field trial test include: (1) long-range deepwater communication for near real-time sensor data collection, and display of key riser health data and analysis of data to meaningful information, (2) advanced machine learning techniques for fatigue damage estimation, and (3) integrated system functionality of the marine riser digital twin system on a drilling riser, which is generic to the offshore drilling industry.
 |
|
Figure 3. Deployment of the sensing element with the subsea sensing module to the clamping element by a ROV. |
Next steps
 |
|
Figure 4. Topside acoustic receiver for data acquisition. |
As the marine riser digital twin advances toward a commercial product, it is envisioned that the entire twin deployment and running process will be handled by standard rig crew teams with no specialty human skill sets required to use the system as intended. When fully deployed, these advanced digital technologies will change the way we work and improve the integrity and performance of our assets. This is just the latest example of GE using software to improve subsea drilling. GE also has partnered with customers to optimize valve assemblies by adding cloud-based analytics running on GE’s Predix digital platform, the operating system for the Industrial Internet.
The machine learning model – the virtual model of the riser is continuously updated against the sensor measurements and metocean conditions, and a model retraining is automatically trigger when the data is beyond the original training conditions. The continuous learning and update of the digital twin model allows us to provide more precise calculation of the riser fatigue life.
Special thanks to Sonardyne and Seanic Ocean Systems for their invaluable contributions to this program.
Greg Myers is a senior product manager within the Subsea and Drilling unit of GE Oil & Gas. Greg has worked in the field of drilling and downhole measurements for his entire career, beginning as a wireline field engineer. Greg studied at Rutgers University and earned a bachelor’s degree in Geology.
Judith Guzzo is a senior research scientist at GE Global Research, focused in the Software Sciences & Analytics and Robotics domain. She has over 20 years of technical and project leadership experience in asset visibility technologies of complex systems with 9 issued US patents and over 15 peer-reviewed journal and conference proceedings.
Shaopeng Liu is a lead scientist at the Software Science and Analytics division of the GE Global Research Center in Niskayuna, New York. Shaopeng has concentrated his focus on the research and development of cyber-physical system technologies and solutions for diverse industrial applications. Shaopeng received his Ph.D. degree in Mechanical Engineering from the University of Connecticut in 2012, and received M.S. and B.S. degrees in Mechanical Engineering from Tsinghua University, Beijing, China, in 2007 and 2004, respectively.
Subscribe
