Safety risk transformation
A probabilistic digital twin may include reliability and degradation models to predict the remaining lifetime of mechanical components. However, it is more than a predictive maintenance tool. Risk is not only about component failures, but also about exposure to hazards and how the asset is operated. A probabilistic digital twin can say something about the overall impact on safety by linking conditions to possible accident scenarios. It can do this by combining reliability models with models of the hazard exposure and the consequences if something goes wrong.
Figure 2 shows the relationship between physical assets, digital twins and probabilistic versions of them. Three main elements distinguish the probabilistic digital twin. First, probabilistic degradation and failure models, reflecting uncertainty and variability of conditions and processes that affect performance and lead to failures. Second, logic and relational models, relating performance variables to failures and loss events. Third, surrogate models, approximating heavier simulation models, allowing fast queries and enabling propagation of uncertainty and model coupling.
Here is a detailed description of a demonstration case involving a probabilistic digital twin monitoring the risk of a burst in a gas export pipeline from an oil and gas platform. A web app demonstration shows how this risk depends on how the pipeline is operated, offering decision support to operators.
Digital twins in the future
Having encouraged the use of digital twins in multiple industries, DNV GL is prepared for a future in which its clients will have them for all their assets. “Companies should not compete on safety, but instead learn and transfer knowledge that keeps the sector safe and secure. We are discussing with several clients how to make their digital twins probabilistic, so that they may be used for risk management,” said Pedersen.
Regulators worldwide require operators to know the status of safety barriers. The challenge is to know how detection of, for example, a degraded component affects the risk level. Operators also want to know if it is safe to continue operating, and for how long, if a problem is discovered. Understanding the safety integrity level can deliver cost savings through better timed and accurate maintenance planning.
DNV GL has developed risk models for several types of safety barriers, in which probabilistic reliability models are updated with information from tests and inspection, providing decision support to the operator. It is actively collaborating with Brazilian oil and gas company Petrobras on this to assess various safety barriers for wells and blowout preventers.
Pedersen said: “Many of our clients are building and maintaining digital twins of their assets. The probabilistic digital twin allows us and our clients to take advantage of all the information such twins contain to improve risk assessments. There is no digital twin that rules them all. Instead, there should be inter-communication between digital duplicates and the creation of a transparent data platform. We must keep pace with this rapidly evolving technology to improve risk assessments.”