The Digital Twin in oil and gas: How far have we come?
We started discussing the idea of Digital Twins in our industries a couple of years ago. When I investigated the popularity of the search term over the last 5 years, it shows that it is now googled more often than ever before.
We started discussing the idea of Digital Twins in our industries a couple of years ago. When I investigated the popularity of the search term over the last 5 years, it shows that it is now googled more often than ever before. To put that into perspective, even today people search the web for the Olsen twins about 20 times more often than for Digital Twins, but still. I imagine the motivation behind both searches could be: “I wonder what they’re up to today”. (The Olsen twins are now fashion designers, so you can skip that search and continue reading this blog post instead.) As for the Digital Twins, what are they up to today?
Hyped for a while
Research and advisory company Gartner listed Digital Twins at the “peak of inflated expectations” in last summer’s Hype Cycle for emerging technologies. Were they right and we’re now walking through the “trough of disillusionment”? Or do we see our expectations fulfilled and the industry benefits from Digital Twin applications in production at scale?
To answer these questions, we need to find out what we actually have in mind when talking about Digital Twins in offshore engineering.
Source: Gartner (August 2018)
Tell me your vision of a Digital Twin and I tell you who you are
Gartner defines the Digital Twin as “a software design pattern that represents a physical object with the objective of understanding the asset’s state, responding to changes, improving business operations and adding value”.
This rather vague description leaves plenty of room for interpretation, and it seems everyone reads it differently. The definition doesn’t say to which extent data analytics is part of the concept (yet for many of us the two topics seem to go hand in hand), and whether Digital Twins are expected to provide hindsight, insight or foresight, for example.
Difficulty and value of various levels of data analytics
Discussing with customers, peers and mentors, I struggle to find the one definition everyone can agree on. For some, any digital model representing a real asset is a Digital Twin. For others, it’s the visualization aspect that adds the value: show me a 3D representation of my ship, offshore platform or wind turbine and I’ll be able to better put operational information into context. While for some, basic information about the “as is” condition of the asset is what makes a digital model a twin, others seek to simulate future and “what if” scenarios as a basis for smarter business decisions.
Digital Twin applications in practice
One group of players in our industries that has widely adopted Digital Twin models into their business are equipment manufacturers, or OEMs. Rotating machinery and other pieces of equipment are being shipped fully instrumented, enabling operational data to be sent in real time. Applying data analytics on these data sets promises a shift from rigid inspection and maintenance regimes towards risk-based approaches, minimizing costly downtime.
This makes sense for manufacturers. Their products are built in smaller or larger series, and the resulting data sets, once well-understood, can be evaluated by the same rules for all units. However, when I talk to operators of oil and gas assets or ships, they don’t think of a pump when they hear the word Digital Twin. They think of a Digital Twin as a representation of an entire production platform or offshore supply vessel, and those systems are huge and unique, one-off solutions. This is not to say the pump isn’t a part of this system, but it raises a new challenge of data integration across multiple sources. Open industry platforms are set to achieve just that, but still have difficult puzzles to solve when it comes to truthfulness of sensor readings, ownership of data sets and compatibility issues.
Focusing our efforts on sub-domains seems promising and more realistic: In Structural Health Monitoring (SHM), the combination of sensor readings from strain gauges, met-ocean data and finite element models allows a more accurate prediction of an asset’s structural as-is condition or even an outlook into its remaining fatigue life. Similarly, a Digital Twin in offshore engineering can be established looking at performance data for topside processes, reservoir simulation etc., but again, these are still unique and complex projects and examples in production today are rare.
Dream big and start small
Making a Digital Twin of the complex assets we deal with in offshore engineering is hard. Even a definition of which capabilities a Digital Twin should have isn’t commonly agreed upon. Going back to Gartner’s definition, maybe we need to look at the last two words, “adding value”, because this very much depends on who you’re talking to. As we saw, a Digital Twin for an OEM might be a clearly defined and manageable project, helping their customers save cost and unlocking new services and revenue streams for the manufacturer. A Digital Twin in oil and gas, on the other hand, specifically for technical authorities and asset managers, might look more like this one from a well-known Science-Fiction series:
One Digital Twin model covering various domains
Inspite (or because) of its Science-Fiction nature, I find this actually helpful! It has everything that makes the vision of a Digital Twin so powerful and complicated, and these are the topics we need to tackle if we want to make this a value-adding reality.
Visualization: All the relevant metrics are displayed on a single 3D model. This is the most powerful way to clearly present aggregated information to decision makers, so that they can make the right calls within short time despite eventual limited in-depth domain expertise.
Data integration: I can see shield strength, power level and torpedo manifest displayed on the Digital Twin. The asset manager needs to have the most important information from across various disciplines at hand. Mapping data across domains to integrate them into one Digital Twin in offshore engineering is not a small task, but it seems to be an important one regardless of whether we’re talking about real-time sensor data, simulated or calculated, or even manually input data.
Collaboration: This Digital Twin is shown on the big screen on the bridge of the (space)ship. That’s because all stakeholders of an asset (or in this case all officers on the bridge) need to work together to make projects and operations a success. We need to cut inefficiencies in how we collaborate and communicate around our engineering models and let Digital Twin applications be the bridge that brings all stakeholders together.
So, we know where we want to be in the year 2245, and we know which building blocks to focus on today to lay the foundation. Let’s get to work and help Digital Twins overtake the Olsens.
Want to know more?
Get in touch with me via LinkedIn or drop a comment below to find out about our most recent project, Sesam Insight. Learn how it helps boosting collaboration in offshore engineering and lets you build shared understanding among all stakeholders through visualization.
Author: Jan Land