Ever since the 1800s when van der Waals first described his Nobel prize winning model for liquids and gases, finding accurate models to describe multiphase behavior has been of significant interest. From describing how hydrogen and oxygen interact in rocket propulsion to complex chemical mixtures, an accurate fluid model, usually an equation of state (EOS), is needed.
Developing a model that can accurately predict 100's or even 1000's of components, where most of the components have unknown properties, using limited data is the challenge that is faced in fluid model development for the energy industry and in CO2 storage projects. On top of the challenge of accurately in describing these complex systems, the industry needs simple models that can run fast. One of the most promising methodologies for providing such models is by first developing a complex detailed model using a cubic EOS and then creating the simple end-point models using this complex "parent" model.
This approach is heavily dependent on the accuracy of the model and whether or not the "parent" model is physical or not. Inconsistencies in the parent model development, due to lack of data, bad data or un-physical tuning, can lead to enormous inaccuracies in the end-point models.
Coupling the different steps of the model development process from sampling, choice of experiments, parent model development and end-point model development with a unified goal of creating consistent and accurate models that can be used by the industry is the goal of this project. This will require breaking the silos between the different stages of the model development resulting in the best possible industry user experience.
The current development procedure for hydrocarbon production and carbon-capture utilization and storage (CCUS) is divided intro three main information silos, namely (1) sampling and PVT experiments, (2) detailed EOS model development and (3) end-point usage of different models (black-oil tables and various lumped compositional models). However, the cross communication between these three segments is limited at best and completely separated at worst. Because each step is dependent on the last (e.g. the detailed fluid model development is highly dependent on the available lab. data), this can lead to uncertainties in the predictability of the various end-point models. This can have a significant impact on economics and safety of the project.
Whitson AS has 30+ years of experience as the connection point between the industry, in developing useful, consistent and accurate models for end-point users. By utilizing this knowledgebase this project has as its goal to bridge the gap between the different silos and develop efficient methodologies for enhancing the fluid model development process. By re-thinking the way fluid models are developed, starting at the beginning of the sampling and PVT data collection and ending with the various models needed by the end-point used, this project aims to interconnect the steps in the fluid development phase in an effective way. The result of this are more consistent and accurate models for the end-point used, reducing uncertainties and risk. The approach to achieving this novel approach of fluid model development is by working with (a) the in-house experts at Whitson AS and (b) to work closely with our industry partners such as operators, PVT laboratories, and academic institutions on each section and try to develop technical bridges between the three segments.