Trustworthy digital twins can power a more efficient future

Published: 23 March 2020


  • Establishing trust will be key to the adoption of digital twins

  • Companies will only be able to extract maximum value from a digital twin if they are assured that it will function as specified

  • A structured, systematic approach is needed to ensure performance matches expectations, including setting clear goals and managing working processes

  • TechnipFMC and DNV are developing the oil and gas industry’s first methodology for qualifying the integrity of digital twin technology


Digital twins provide virtual representations of systems and physical assets over their lifecycles. Recognizing the potential to increase safety and efficiency, the oil and gas industry is increasingly applying this technology. As the industry prioritizes digital twins and technologies integrated within them for investment in 2020, this is set to grow, according to DNV’s industry outlook, New Directions, Complex Choices. The report assesses sentiment, confidence, and priorities for 2020, among more than 1,000 senior oil and gas professionals.

The attraction is that a digital twin can support information-based decisions across the lifecycle of assets, from the design stage through to decommissioning. The big question though, is whether the information from a digital twin can be trusted. Establishing such trust will be key to adoption, as companies will only be able to extract maximum value from a digital twin if they are assured that it will function as specified.

As more digital twins enter the oil and gas sector, it is key for operators to know that their twin works as planned, and that its output is reliable

  • Julie Cranga, vice president, subsea digital, TechnipFMC

“As more digital twins enter the oil and gas sector, it is key for operators to know that their twin works as planned, and that its output is reliable” said Julie Cranga, vice president, subsea digital, TechnipFMC.

TechnipFMC – a global leader in subsea, onshore/offshore and surface projects – is partnering with DNV to develop the oil and gas industry’s first methodology for qualifying the integrity of digital twin technology. It is being piloted in 2020 in two subsea development projects.

To ensure that the performance of digital twins matches expectations, organizations involved require a structured, systematic approach. This needs to provide evidence that the twin will provide valid system information and accurately predict system performance. To be qualified under this approach, a twin would need to achieve these tasks within well-defined limits and to a stated level of confidence, and to do so over time.

Kjell Eriksson, regional manager for Norway, DNV GL - Oil & Gas

Clearly specifying what you want to achieve from a digital twin and how you will manage the change or disruption to working processes is a genuine challenge for many companies

  • Kjell Eriksson, vice president – digital partnering, DNV

“Bringing clarity and structure to the qualification process is critical,” said Kjell Eriksson, vice president – digital partnering, DNV. “In these early years of the industry using this technology, clearly specifying what you want to achieve from the digital twin and how you will manage the change or disruption to working is a genuine challenge for many companies.”

Overcoming specification and work process challenges

To help overcome these challenges, the industry will benefit from considering twins in terms of functional elements. A functional element is the part of a twin’s capabilities that supports the user in making a specific decision. For example, the process and checklist for a functional element should cover: the need that it must meet; the key decisions it must support; and, the types, quality and number of data sources such as sensors that will be required. The feasibility of the functional element will be continuously assessed as it matures and is ultimately validated. The functional element should also cover the performance of the digital twin over time, as expressed through a quality indicator.

The quality indicator reflects the fact that users need to be able to trust the digital twin at any and all times. For example, a sensor may have stopped working without anyone noticing, or maintenance and modification work on a physical asset may not have been updated in the virtual model. 

“Through a structured combination of automated and audit-based assurance processes, companies can trust that the twin is always up to date,” said Eriksson.

A scalable and broadly accepted methodology will benefit all parties

A methodology to qualify digital twins should adhere to the following three principles (Figure 1).

First, it must enable a modern agile approach to development of digital twins, while being systematic at the same time. Second, it should be scalable as digital twins mature and evolve. Finally, the qualification process must become broadly accepted in order to serve as terms of reference between supplier and buyer, enabling efficient development and procurement processes.

The methodology that TechnipFMC and DNV are developing aims to meet all three of these goals.

Data and algorithm quality is vital in digital twins

Advanced industrial operations depend on information systems for control and analysis. Data is increasingly considered to be of equal value to physical assets, and considerable costs are involved in collecting, storing, and acting upon data. An advanced digital twin is no exception.

The quality indicator component involves continuous assessment of the quality of data and algorithms, and of the twin’s output and automated recommendations. The indicator also includes periodic assessment of the functional elements and of changes to the asset.

The methodology and logic for how to ensure quality of data and algorithms is based on two recommended practices: DNV-RP-0497 Data quality assessment framework; and the forthcoming DNV-RP-0510 Framework for assurance of data-driven algorithms and models. DNV-RP-0497 includes a process for organizational data maturity assessment.

A recommended practice to assure digital twins

DNV’s Technology Outlook 2030 forecasts the emergence this decade of a full digital value chain in oil and gas, with the digital twin at its core, to reduce development times and costs in the energy transition.

As the TechnipFMC/DNV pilot project continues, the aim is to refine the methodology to publish by the end of 2020 a new DNV recommended practice for the quality assurance of digital twins to increase their efficiency.

While the methodology is a first for the oil and gas industry, it is being built on tried and tested foundations.

It is derived from Recommended Practice DNV-RP-A203 Technology qualification, and from other standards, adapting them for use on digital twins. DNV-RP-A203 was first published more than 20 years ago as a common framework for oil and gas industry players to gain acceptance for implementing unproven hardware technology. It has been used to demonstrate the trustworthiness of hundreds of technologies.

The digital twin qualification methodology being developed uses the definition of digital twins previously provided. It is applicable to alternative definitions, though this may require adjustments depending on the scope and application at hand.

“We invite other companies to come forward with digital-twin modules with which we can test and refine the assurance methodology before the recommended practice is published,” said Eriksson. “We aim to create a broadly accepted recommended practice that operators and technology providers can use as a key reference. Having a standardized approach will remove uncertainty and move the industry to a more efficient future powered by trustworthy digital twins.”

Related links

  New digital twin concept could show real-time status of safety risk and operations

New digital twin concept could show real-time status of safety risk and operations

Read here

  Machine learning can make mooring safer and more cost effective

Machine learning can make mooring safer and more cost effective

Read here

  DNV Data quality services

DNV Data quality services

Download brochure here

  Machine learning assurance

Machine learning assurance

Read here

Sign up to receive PERSPECTIVES

 

Digital publication from DNV

Sign up to receive 'PERSPECTIVES' - a digital publication offering insights from our people, customers, and industry colleagues.

Disclaimer: 

DNV prides itself on providing accurate information but makes no claims or guarantees about the accuracy, completeness or adequacy of contents in this publication, and disclaims liability for any errors or omissions. The authors’ views here do not necessarily reflect DNV’s views.