Abstract
Due to the global increase in energy demand, the need for economic oil and gas production is rising more than ever. Therefore, it is necessary to ensure that subsea architecture designs are economical and safety oriented. While numerous challenges are encountered during subsea system’s installation and operation phases, most of these challenges can be avoided by ensuring an economical and reliable design. For a safe and cost-effective design and operating scenario, it is essential to predict the hydraulic and thermal behavior of multiphase fluid encountered in petroleum pipelines for a range of conditions. This cannot be accomplished by empirical models, which are dependent on limited data available. Consequently, mechanistic low-dimensional models have been used for two-phase gas-liquid steady-state flow. However, mechanistic low-dimensional models assume adiabatic conditions, which is rarely the case in subsea architectures, which encounter cold surroundings. Therefore, to predict the temperature-based characteristics of multiphase flow in environments with thermal gradients, a thermal model has been developed and validated with experimental data. 80% of the validation data was predicted by this developed thermal model with error difference of less than 30%. The developed two-phase gasliquid thermal model was merged with Beggs and Brill hydraulic multiphase flow model to predict the overall behavior of two-phase gas-liquid flow, and used to develop an optimal model-based multi-well subsea architecture design. A case study of a four-well subsea system was used to demonstrate the automated subsea architecture optimization technique. Through this case study, it was shown that approximately 23% of savings in pipelines procurement could be made relative to the conventional designing approach. Industry standards, safety factors, and multiphase flow models were used to design jumpers and place the manifold for a subsea multi-well system. Merging hydraulic and thermal multiphase flow models showed the effect of temperature on the flow, which led to an optimized design for the subsea insulation in which issues such as wax deposition can be prevented. The resulting optimized subsea architecture was then implemented in Simscape® environment to obtain the transient response. Along with optimized subsea architecture automated design, the developed thermal model has the potential to be used for real-time prediction of two-phase flow rate, pressure drop and void fraction as virtual sensors to provide economical alternative to expensive and impractical hardware sensors. Furthermore, the developed model can also be used to design effective control strategies for multiphase flow regulation in jumpers and prevention of backflow at the manifold.
Ensemble Machine Learning Assisted Reservoir Characterization Using Field Production Data–An Offshore Field Case Study
