1. SUBSEA DEHYDRATION PROCESS CONCEPT AND INDUSTRIAL OBJECTIVES
The main objective of this project is to design, optimize and introduce a reliable membrane-based separation process for the subsea dehydration. The process includes individual membrane systems working at different operating conditions in a closed loop, where TEG is the absorbent agent in the loop (Figure 1). This work benefits the obtained experimental data from each membrane process to optimize the process according to the required specification provided by our industry partners (i.e. Equinor). The main criteria for the process are to meet the specification by lowering the water content (dew point < -18°C at 70 bara) and TEG content (TEG emission < 0.8 liter/MSm3 gas) in the gas pipelines.

2. MEMBRANE PERFORMANCE TESTING AND EXPERIMENTS
Durability of membranes and the TEG flux through the membrane will be evaluated experimentally to ensure longer lifetime of membrane materials. Membranes will be tested at different transmembrane pressure difference to optimize the safest pressure difference across the membrane. Membranes with longer lifetime such as hydrophobic polymeric membranes and inorganic membranes will be produced and tested in membrane contactor and membrane
thermopervaporation for higher flux and separation performance.

3. MULTICOMPONENT MODEL DEVELOPMENT AND VALIDATION, PROCESS DESIGN, AND OPTIMIZATION
A mathematical model describing transport properties of components will be developed in Python and then be implemented in Hysys software. Different configurations in the model (1D-1D or 1D-2D), membrane/module geometries (hollow fiber, flat sheet) and flow configurations (co- or counter current) will be evaluated to find the optimized configuration. Furthermore, to address all the gas components in the model, novel thermodynamic correlations are required to be defined. Accordingly, a multicomponent mathematical model with fast and reliable solvers will be employed for solving the equations in the simulator to optimize the required membrane area and the cost with the smallest footprint. At the current state, 1D-1D models for water concentration and temperature profile were developed for membrane contactor in Python and the preliminary evaluation of the models show low water content in the gas outlet in the scale of laboratory membrane module and flow (Figure 2(a-c)). The model will be verified
first with data obtained from laboratory and then be implemented in Hysys to predict the performance for larger membrane module with larger packing density. In a later stage, the model for thermopervaporation unit similar to membrane contactor will be extended and implemented in Python and Hysys to minimize the energy and cost for the process while enhancing the regeneration performance.
With subsea dehydration, the design will be validated and the water content in gas stream is expected to be in the rage of pipeline specification with small footprint and less complexity. The obtained water from the gas phase is expected to be pure (>99.99 %) to be discharged to the sea water safely.