Hollow-fiber membrane contactors for CO2 capture: design of a hydrodynamically optimized process

Offer Description

General context

Increasing emissions of greenhouse gases, of which carbon dioxide (CO2 ) accounts for around 65% (Bains, P. in 2017), contribute massively to global warming. In order to limit these emissions, it is necessary to capture the CO2 emitted by industrial and energy sources (coal-fired power plants, cement plants, refineries and chemical plants). The CO2 captured is then stored or used as a raw material for the chemical industry, for example in the production of methanol.

In post-combustion flue gases, industrial CO2 emissions are not very concentrated and occur at atmospheric pressure. In fact, CO2  generally represents less than 15% of the volume of these flue gases, which are also composed of oxygen, water vapor and nitrogen. It is therefore necessary to separate the CO2 from the other compounds. Several separation technologies exist, among which absorption in chemical or physical solvents plays a predominant role. This operation is generally carried out in packed columns. As part of an intensification approach, hollow fiber membrane contactors (HFMCs) offer an alternative to columns, particularly for physical absorption, characterized by particularly low liquid-side transfer coefficients and particularly high liquid-to-gas mass flow ratios (Bazhenov, D. in 2018). In HFMCs, one phase flows inside a large number of small-diameter tubes (the hollow fibers) arranged in parallel, and the other phase flows around the fibers, in the shell, Fig. 1. The fiber wall is usually covered with a dense polymer membrane permeable to CO2 (and often not very permeable to other compounds). The specific interfacial area available for transfer is much higher in HFMCs than in packed columns, potentially leading to more compact systems for the same transferred flux.

Objectives of the PhD

Today, the design of these contactors is based on the strong assumption of ideal flow (non-dispersive piston) inside and outside the fibers. This assumption is rarely verified in practice, since maldistribution of flows is often observed, resulting in a decrease of separation efficiency. In fact, the geometry of the distributors and the non-uniformity of the diameter and arrangement of the fibers lead to preferential passages, reducing the efficiency of CO2 removal. Adequate characterization of the hydrodynamics is therefore necessary for more reliable sizing of such equipment.

The first objective of this thesis project is to study flow maldistribution in HFMCs. This will be followed by the development of geometries with optimal distribution. The geometries developed will then be integrated into a multi-stage absorption process, the architecture of which will be optimized to maximize performance. To achieve these objectives, a dual numerical and experimental approach will be adopted.

Scientific methodology

First, a literature review will be carried out, followed by a numerical CFD study (using ANSYS Fluent software) of the hydrodynamics of gas and liquid flows in these contactors. In particular, two-dimensional simulations will be carried out with a porous medium representing all the fibers. The velocity field will be calculated by solving the Navier-Stokes or Darcy equations, coupled with the mass conservation equations. The residence time distribution (RTD) will be calculated by simulating the transport of a passive scalar injected at the inlet, which concentration will be recorded at the contactor outlet. The RTD will be used to analyze the links between operating and geometric conditions, on the one hand, and distribution quality, on the other. The consequences of maldistribution on CO2 separation performance will then be assessed.

In parallel, an experimental bench will be developed at the LRGP. Pressure drops in the gas and liquid phases will be measured and compared with the values obtained by numerical simulation. Residence time distribution measurements will be carried out, using ozone as a tracer in the gas (poorly soluble in water) and salts (NaCl) in the liquid. An ozone generator will be connected to the gas inlet, and a UV analyzer will monitor the ozone concentration at the outlet. At the liquid outlet, a conductivity meter will track the salt concentration. RTD analysis will enable the detection of dispersion, dead zones and possible short circuits. In a second phase, experiments to visualize the flow in the shell will also be carried out.

Existing membrane contactor geometries will first be considered, then new geometries will be developed with the aim of limiting maldistribution. Several geometrical parameters (distributor shape, position of inlets and outlets, etc.) will be explored and optimized using the optimization module of the ANSYS Fluent software.

The most efficient geometries will be integrated into a multi-stage physical absorption process coupled with a desorption module. To handle high flow rates, membrane contactors will be arranged in parallel within modules which shape will be optimized to provide an even flow distribution and a maximum performance. Several modules will be placed in series to maximize absorption. Process architecture and operating conditions will be defined and optimized by process design calculations, taking into account the different scales of the problem.

Presentation of the host laboratory

This thesis will be carried out at the Laboratoire Réactions et Génie des Procédés (LRGP, UMR7274 CNRS-Université de Lorraine, Nancy), which employs 300 people working to develop the scientific and technological knowledge needed to design, study, manage and optimize complex physico-chemical and biological processes to transform matter and energy.

The PhD student will be based in the PRIMO team, which focuses on the design, study and optimal operation of new processes involving intensified reactors and microstructured systems, membrane processes, gas/liquid contactors and supercritical processes.

The thesis is part of the ANR project IMOSYCCA (Intensified Modular Systems for friendly CO2 capture), which involves several French laboratories including LRGP, and the PEPR SPLEEN program (Priority Research Program and Equipment "Supporting innovation to develop new, largely carbon-free industrial processes").

Funding category: Contrat doctoral
Net salary : 1715.89 € euros /month for the research activity, which may be completed by a paid teaching activity
PHD title: PhD of Process and Product Engineering
PHD Country: France

Requirements

Specific Requirements

Candidates profile

The candidate will be an engineer or hold a Master's degree in process engineering or fluid mechanics. He/she is interested in both numerical and experimental work, and should have a good skills in several of the following thematics: fluid mechanics, transfer phenomena, numerical methods, process design, separation processes, residence time distribution. Knowledge of computational fluid dynamics (CFD) would be an advantage.

Student Profile

We are looking for a highly-motivated PhD candidate, of Master’s level in process engineering or fluid mechanics who likes experiments, modelling and numerical simulation. More specifically:

Prerequisites

Very good level of English, spoken and written.
Knowledge of process engineering, in particular separation processes
Knowledge of fluid mechanics and transport phenomena.
Experience in Computational Fluid Dynamics simulations
Experience with experimental techniques