Next generation flow assurance models for wells and risers in LedaFlow

The industry is at present operating only with rather approximate replicates of what is observed in the field for vertical flow lines. Not properly understanding these flow phenomena results in poor designs that in turn lead to significant operational difficulties. This is both costly and it causes potential safety issues. LedaFlow (TM) is a commercially available multiphase flow simulator which has been developed based on models that represent the actual physics of multiphase flow. This project will generate new and validated functionalities and will open a new area of application (namely predictions for wells and risers) in the multiphase simulator LedaFlow (TM). The accuracy of the predictions will be significantly increased by combining: - The unique new large-scale experiments at industrial relevant diameter, length, angle and pressure requirements proposed as part of this project, - A team of experts, able to develop models for the more complex multiphase vertical flow phenomena. - Implementation in LedaFlow (TM) through modelling of the physical multiphase phenomena. The ability to accurately predict high angle multiphase flow in wells and risers will be a strong market differentiator for LedaFlow (TM). It represents significant value creation for LedaFlow Technologies DA (LFTDA). As part of the project, we have successfully performed a unique experimental campaign at the SINTEF Multiphase Flow Laboratory at Tiller. We have obtained many data for vertical flow at relevant industrial conditions in terms of dimensions and pressure. Using this large pool of large-scale validation data, we are now improving several of the models for vertical flow. For instance, we understand now better the interaction between the gas and the liquid in the well, and also how small amounts of condensated liquid accumulates in the well leading to reduced or terminated production. This new knowledge was implemented in the software update release of LedaFlow (TM) (LedaFlow 2.0, April 2016). The end users of the software have already expressed the high value of the model improvements as they enhance the reliability of several predictions and hence the profit of subsequent operations. The project work in 2017 has focus on expanding the acquired knowledge and new models to highly viscous fluids. We have also worked on the development, implementation and testing of an improved droplet entrainment model. The results show improved predictions. Further, the changes in the numerical procedure have also led to increased computational speed. These improvements are now implemented in the commercial software. In 2016 we looked at how the friction factor and the pressure losses in a well are affected when the inner pipe is not concentric with the outer pipe. This is known as "Asymmetric annulus". We have implemented pragmatic corrections that are in line with the physics while they still don't diminish the computational speed. Further, the proposed correction takes into account the type of fluid (Newtonian or non-Newtonian) and also if the flow is laminar, turbulent or in the transition zone. This work has been implemented in the latest software update release of LedaFlow (TM) (LedaFlow 2.1, September 2016). The experiments performed in the project have also provided new knowledge about the "churn flow" regime - a common but little understood chaotic flow phenomenon that can occur when gas and oil are transported in the same vertical piping. The fresh measurements have provided more knowledge about how the chaotic nature of the phenomenon affects the ability of the well pressure to push the gas and liquid further through the multiphase pipelines. A new functionality has been developed focusing on "churn flow". In 2017 we have tested exhaustively this new functionality and it is now implemented in the commercial software as well. Finally, we have made improvements in understanding and properly modelling the transient behaviour in pipe flow with high angles or vertical flow, which is usually simulated using the "Slug Capturing" approach in LedaFlow. The objective of this work is to obtain better predictions in dynamic flow situations, in particular for slug flow, which is a flow regime that can cause significant operational problems if not accounted for in the process design. For instance, the intermittent liquid rates associated with slug flow can cause mechanical failure in spools or flooding of the separation facilities, leading to production loss and possibly danger to personnel. By improving the Slug Capturing model in LedaFlow, the transient flow characteristics are now more accurately predicted, allowing for more optimal design and operation of slugging transport systems.

This project will generate new and validated functionalities in the transient multiphase simulator LedaFlow(TM) in a new area of application: flow assurance predictions for well and risers. - Reliable predictions of pressure loss and liquid content in vertical flows under stationary conditions for all types of flow regimes and under transitions. Understanding and subsequent modeling of churn flow will be given special attention. - Accurate predictions of the onset of liquid accumulation in wells, thereby expanding their lifetime. - Realistic predictions of highly transient/developing flows in wells and risers. Current vertical flow models are largely based on rough correlations which are not expected to scale well beyond their calibration domain. Consequently, these models are expected to be inaccurate under conditions that deviate from available validation data. A more physically correct model is believed to provide more accurate predictions and better scalability. In order to achieve the above, this project includes a unique experimental campaign at relevant industrial conditions: large pipe diameters and high pressure. At present, no simulation tool has access to such a pool of large-scale validation data. By having this access, LedaFlow(TM)will offer far better scalability and accuracy than any other commercial software in the market. New models in LedaFlow(TM) that are based on laboratory data from industrial scale and conditions will give extremely valuable operational knowledge, as well as ensure a reduced uncertainty in flow assurance predictions. The end users will benefit from this project as it will allow for a significant reduction of risk and for cost effective design and operation of wells and risers. The project objective fits very well with the top priority in OG21-TTA4:" Fundamental knowledge on multiphase pipeline flow and flow assurance"