Aalborg Universitets testcenter

Milestone for artificial intelligence within control of oil and gas production

Thursday 04 Jan 18
|

Collaboration between researchers and industry should increase production and prevent human errors, for example in handling of water in oil production in the North Sea. 

The oil produced in the North Sea contains an increasingly large quantity of water because oil reservoirs are gradually filled with water as they are emptied for oil. The water must be handled so that it remains clean enough to be discharged again, to either the sea or underground. Also, large volumes of water are pumped up from the sea and treated before it is injected to reservoirs for pressure support to increase the oil production. These processes related to water handling are complex and they involve large volumes of data, signals, values, and alarms requiring extensive monitoring by control room operators.

To ensure a process in which all situations can be handled as early as possible without impacting production and without risk of human errors, a new broad collaboration has been initiated between researchers and companies on automation of the monitoring. The collaboration comprises DTU, Aalborg University, Centre for Oil and Gas - DTU, Maersk Oil, ConocoPhillips (UK) Limited and the Norwegian company Eldor Technology. In addition, the joint project receives funding from the Research Council of Norway, Centre for Oil and Gas - DTU and Innovation Norway.

“At a recent demonstration, we’ve shown that our idea of achieving further automation using a functional model, a so-called MFM model, combined with large data volumes and artificial intelligence, is both possible and will improve the existing monitoring. We’ve thus reached a large and important milestone, not only in the existing project with oil and gas production monitoring, but also in our ability to use the model for automatic monitoring in all other industries,” says Professor Emeritus Morten Lind, DTU Electrical Engineering.

The idea has been clear from the outset, but it has required interdisciplinary efforts with contributions from all collaboration partners to be able to prove that the idea can also be deployed in practice as a decision support tool for the control room operators.

MFM modelling developed at DTU
The MFM modelling technique has been developed by Morten Lind based on his research into the interaction between humans and machines over the past 30 years. The point of departure for the research is that, by means of data and knowledge about the processes behind, for example, oil recovery—combined with insight into the laws of physics—a monitoring system can be developed which automatically detects any abnormal situation, highlights the root cause, anticipates future consequences, and provides instructions on how to rectify the situation.

The MFM technology has been developed based on the need to control and monitor complex energy systems and industrial processes and can be used widely in process and automation design, as well as for operational and maintenance support. MFM can therefore serve as the technological base which integrates and share knowledge between several phases in the life cycle of a processing plant.

“The specific task of handling water treatment systems in oil and gas production is thus just one application example. The MFM technology is suitable for many other complex industrial processes, in oil and gas production, and equally also for automatic control and monitoring systems in completely different industries,” says Morten Lind.

The current work to feed the model with the correct data and to adapt it to the existing oil and gas control systems has been performed utilizing expertise from Centre for Oil and Gas - DTU, DTU Electrical Engineering, Eldor Technology, as well as Aalborg University. Aalborg University’s test centre in Esbjerg—which comprises an oil production model—is also the location at which the new decision support system with the work name AlarmTracker was tested.

Researchers from both DTU and Aalborg University are thrilled to participate in the development of the new control system.

"The development of AlarmTracker has become even more relevant than when we started the project three years ago. Both the development in oil prices and the great focus on digitization in all industries mean that there is an increasing need for better monitoring of industrialized processes. "
Managing Director Bjarne André Asheim, Eldor Technology

“It’s highly motivating. All partners in the project are concurrently working on the development of a concrete product, which must be completed in the course of a relatively short time. We are a piece in a joint puzzle in which we need input from the other participants, and we can see how they use our contributions to proceed with their part of the project. We’re all dependent on each other, to enable us jointly to create a new product”, says Research Assistant Stefan Jespersen, Aalborg University.

Automation is key in unexpected events
So far, control room operators have had to monitor a wealth of information in connection with, for example, the handling of water, both from the production and the water injected to increase oil production. So much information can be difficult to grasp when something unexpected suddenly happens. Automation will facilitate this task by providing information to the operators about what is wrong, what the consequence of the fault will be, and what measures are to be taken to prevent it.

The most important contributor to software development in the project is the company Eldor Technology, which is looking forward to the next demonstration in the spring. Here, the system will be demonstrated using real-time data, integrated with the HMI (Human Machine Interface) systems, which are currently used for monitoring.

“The development of AlarmTracker has become even more relevant than when we started the project three years ago. Both the development in oil prices and the great focus on digitization in all industries mean that there is an increasing need for better monitoring of industrialized processes. The project is progressing according to plan, and we therefore expect to be able to launch our new product in about a year from now, at the beginning of 2019”, says Managing Director Bjarne André Asheim, Eldor Technology.

When the new product is ready, the buyers will also be in place. Centre for Oil and Gas - DTU has entered into a contract with Maersk on the project, and the company Eldor Technology collaborates with the international oil and gas company ConocoPhillips (UK) Limited.

After the demonstration, there was, in fact, no doubt among the invited guests from the two companies.

“We see many exiting perspectives and are very positive about the AlarmTracker project” said Team Lead E&I Torben Ipsen Bruun from Maersk Oil and Dale Allison from ConocoPhillips (UK) Limited said AlarmTracker intends to enhance the situational awareness of control room operators and we are excited to receive the predicted benefits when developed” after the demonstration.

More information:
Erik Bek-Pedersen, Programme Manager, Centre for Oil and Gas - DTU, email: erikbp@dtu.dk, tel.: +45 6114 0732 

Bjarne André Asheim, Eldor Technology, email: bjarne@eldor.no, tel.: +47 95 29 14 78

 

PhotoTest plant at Aalborg University, at which a recent demonstration has shown that the project’s idea of achieving further automation using a functional model, a so-called MFM model, combined with large data volumes and artificial intelligence, is both possible and will improve the existing monitoring.

News and filters

Get updated on news that match your filter.