What’s in a Digital Twin?
But before the current trend, we had something we called “Asset Model” or “Data Models” and it was really powerful. My colleague Murray described it better when he wrote:
“Every Digital Twin needs a data model somewhere inside it. The Data Model is the digital map of where everything is; it’s the central reference for how everything is connected; it’s the search engine for your information.”[1]
The Dubai Petroleum Success Story
Matrikon, one of the pioneers in process data visualisation technology, had a technology called NetObjects, which could be configured as a data model of the assets that belonged to an operation. Coupled with ProcessNet (which was many years ahead of its time for web-based visualisation) and a powerful back-end calculation engine we developed Well Performance Monitor (WPM) which included a simple but standardised asset model to represent wellbores, completions, reservoirs and the associated data and instrumentation for the subsurface and surface.
Back in 2009, this asset model enabled the deployment of a field-wide well performance monitoring system for the 300-well Dubai Petroleum operation distributed across five fields. Very few standard templated displays and trends were used to monitor all the producer and injector wells, because the asset model “knew” what tags to use in the case of each well. The configuration and deployment was simplified which minimised the overall project duration and cost and the Dubai-based engineers were delighted to be able to monitor their gas lifted wells from the comfort of a web browser.
The customer was so pleased with the result that decided to publish an SPE paper with numerous examples of the gas lifting anomalies detected and analysed through the use of WPM[2]. In the abstract of the paper, we can read:
“The system has helped identify well performance-related issues in a very short period of time and aided in solving these problems with less production downtime than before. The tool has helped reduce the time spent by engineers for data gathering and issue identification. As the system is web-enabled and deployed through the company’s Intranet, it is possible to monitor the live well performance from any location.”
There are a lot of hard benefits in that short paragraph!
I remember visiting the Dubai Petroleum office after the deployment and seeing all the engineers “chilling out” in the knowledge that production was under control. Quite different when compared with the production loss firefighting of months before.
Most importantly we learned a key lesson: This asset model did not follow any of the self-proclaimed universal ontologies out there, it did not have to include minutia objects or attributes that nobody cared about to support the functions of the application and deliver value: The model could perfectly be fit for the purpose of the use case it needs to support and solve.
Now, NetObjects was a proprietary technology, the data in the model was in a sense “locked in” within Matrikon solutions and software. It also used conventional database technology, with constraints in the way it represented relationships.
The Benefits of Today’s Open Asset Models
Fast forward 10 years, we still use the same Asset Model concept, but implemented on technology that provides even more benefits.
Eigen Ingenuity’s Digital Twin includes an asset model built on Neo4j, an open-source graph database technology used by blue chip companies around the world to map relationships between entities and drive business logic and insights.
A view of an Oil Field Asset Model using the Neo4j web interface.
So, what does this mean for you as an Oil & Gas engineer?
- You can get going from day one. Use cases can be supported one at a time: You can start building your most important workflows first because the model can grow organically to support more and more use cases over time. This enables an agile, organic growth of functionality and benefits.
- You are not locked in. The model of your assets is not locked-in any proprietary technology. The Neo4j asset model is open and accessible by third party applications within your application landscape. Other applications can take advantage of the relationships built in Neo4j.
- It’s Flexible. A graph database is non-hierarchical, which means that different users can view the model from different angles: A General Manager may be interested in production by fields and areas; a Reservoir engineer may be interested in production by reservoirs or zones; a Maintenance Engineer will only be interested in equipment hierarchies: All of hierarchies can be represented and co-exist within a graph database.
- It Performs. It’s fast because it is based on the latest cloud-native and web IT technologies which are under continuous development and improvements.
At the heart of every Digital Twin is an asset model: An old concept, by IT timeframes, but still very useful, if the focus is always on the use case and the value that engineers want to derive.
[1] https://eigen.co/what-makes-a-good-datamodel/
[2] https://onepetro.org/SPENATC/proceedings-abstract/10NATC/All-10NATC/SPE-126682-MS/108581
