SUM - Scaling and Uncertainty Modelling in multiphase production

Multiphase flow modelling is one of several important milestones in the history of Norwegian petroleum development. The technology led to considerable savings in field development, in addition to enabling tie-ins from subsea developed satellite fields, with multiphase transport to an oil platform or to shore. The work with development of the multiphase flow simulator(s) began in 1980 and was originally based on a mathematical model of the cooling cycle of IFE?s nuclear reactor in Halden. Modern multiphase flow simulators depend on tuning and verification with data from field or laboratories. Data from laboratories are often subject to debate given the large difference in scale between the laboratories and the field. Producing relevant laboratory data is often prohibitively expensive which leads to most experiments being conducted with simple model fluids such as Diesel or Kerosene with Nitrogen or SF6 as the gas phase. The SUM-project has established mathematical scaling laws and shown that it is possible to perform relevant and scale-able experiments on a small scale, if a few assumptions are met. This implies that relevant experiments now are possible to perform with considerably reduced cost for the industry. Within the field of multiphase modelling, little effort has been placed on quantifying uncertainty in the calculations, which include numerical errors, effects of scale, experimental errors and human factors, to mention a few. The SUM-project has developed mathematical models for uncertainty in multiphase flow simulators. The results will potentially reduce risk and thereby cost for future field developments. The project has been a collaboration between IFE, SINTEF and NTNU, as well as industry partners Equinor, Lundin, Eni/Vår Energi, Gassco, Ledaflow, Schlumberger, TechnipFMC and the Research Council of Norway.

Methods for uncertainty quantification have been studied 1) through experiments, 2) from experimental data into models and 3) through modelling in multiphase flow applications. A postdoctoral candidate has performed experiments, using a defined scaling principle for multiphase flow and compared results with experimental data from different scales. The scaling principle is improved defining the necessary variables to match for a good comparison. IFE and SINTEF have contributed with their knowledge on the development of advanced engineering models for multiphase flows and investigated further the concern of uncertainty. NTNU's group for biomechanics have contributed with complementary competence on uncertainty quantification for mathematical models through the PhD candidate. The project results demonstrate the necessity of understanding the effect of the uncertainty and their propagation in the models, and how specific variables in multiphase flow have more impact than others.

The project application SUM covers two important and unresolved topics in multiphase flows; scaling and uncertainty. Cooperation between IFE, SINTEF and NTNU will provide a new combination of competence for increasing the confidence in multiphase flow modelling, looking more thoroughly into these topics. Scaling and uncertainty have been in focus in internal projects in the institutes individually, and it is the first time they are put together in a larger context. By assembling existing and new multiphase flow data from laboratories at different scale, we will improve the understanding of specific parameters in scaling rules. Previous studies have shown the importance of matching density ratio for scaling purposes, and new studies may establish knowledge of the importance of other parameters as the interfacial tension. Even if uncertainty quantification in computations and simulation tools has surged over the last 10-15 years, little work has been done in the context of multiphase flow in pipelines. We aim here to define a systematic methodology for assessing uncertainties in 1D simulation tools for multiphase pipe flow, with focus on scaling effects. IFE and SINTEF are world leading contributors to fundamental and applied knowledge of multiphase flows, and have a very successful record in the development of advanced engineering models for multiphase flow in pipelines and wells. Both institutes have world leading multiphase flow laboratories that have been greatly improved through the National Infrastructure of Multiphase Flow. This project is therefore well aligned with the strategies of the institutes, giving a unique opportunity to improve the model activities involving the important topics scaling and uncertainty.