|Natural gas in the reservoir is saturated with water, which may condense during transportation causing flow assurance problems such as hydrate formation and corrosion. Dehydration of the natural gas is one of the main processing steps in natural gas treatment as removal of the water prevents these transportation challenges. Moving the dehydration process subsea closer to the well gives possible advantages, such as no need for continuously injection of chemicals to prevent hydrate formation. In addition, subsea dehydration enables directly export of the treated natural gas to the export pipelines without the need for further topside treatment. Membranes are considered as a potential suited technology for subsea natural gas dehydration as it meets the subsea design requirements, with high modularity, high flexibility and compact design.
As a sub-project in the SUBPRO research center, the objective of the present work is to perform a first evaluation of a subsea natural gas dehydration process with the use of membrane processes through modelling and process simulation. A new proposed dehydration process is evaluated with the use of a membrane contactor with triethylene glycol (TEG) for dehydration of the natural gas, in combination with thermopervaporation for regeneration of the TEG. Aspen HYSYS V8.6 is used to perform process simulations and optimization of the proposed process, but the simulation software does not have unit operations for the membrane units. Therefore, models are developed for both the membrane units and implemented into Aspen HYSYS with the use of MATLAB CAPE-OPEN.
The membrane contactor model is based on a hollow fiber module configuration with a 1D-2D modelling approach. The model calculates temperature, concentration and pressure changes along one fiber of the membrane contactor. To adjust for the high pressure subsea operation a new fugacity coefficient model is developed for the vapor-liquid equilibrium model, used to calculate the mass transport over the membrane. The developed membrane contactor model is validated against high pressure experimental data, indicating that the model is conservative with a good prediction. The modelling results reveal that to ensure long term stable operation subsea the use of a dense membrane layer on top of the porous support would be preferred to prevent possibly wetting of the membrane pores which significantly reduces the separation performance.
The thermopervaporation model is based on a plate-and-frame module configuration with repeating channels for the feed solution, the air gap and the cooling water. A temperature dependent permeability correlation for the dense membrane layer is developed based on experimental results and included in the thermopervaporation model. The developed model can predict the separation performance, in addition to being used to evaluate the effect of different operation conditions and design parameters. Temperature drop of the liquid feed, due to evaporation and heat transfer between the hot and cold liquid, is found to be a limiting factor as it reduced the driving force and the separation performance. The air gap is found to be a critical parameter in the design of the thermopervaporation module as it acts as an insulating layer for heat transfer.
Three different process designs with respect to staging of the regeneration with heating between each stage are considered in the process optimization, with one, two or three stages. The optimization variables for the systems are the number of fibers in the membrane contactor, the number of feed channels in the thermopervaporation unit, and the flow rate of TEG. The results indicate that the temperature drop of the liquid feed is a limiting factor and that staging of the regeneration is preferred as it reduces both the sizes of the membrane units, the TEG flow rate and the energy demands. The plate-and-frame module configuration of the thermopervaporation unit provides a low packing density, which gives large membrane volume.
To conclude, membrane contactors show a promising potential for subsea natural gas dehydration and thermopervaporation for the regeneration of TEG. The proposed process is able to dehydration the natural gas to the pipeline specification. However, the liquid temperature drop in the thermopervaporation unit and the low packing density of the plate-and-frame module configuration are evaluated as potential limitations for the proposed process design.