Simulation-based prediction of performance-defining parameters of forward osmosis (Working Title)
Background MionTec
Among the pressureless membrane processes, forward osmosis has been gaining more and more importance in recent years, characterized by a steadily growing number of publications. While competing, likewise pressureless membrane processes such as membrane distillation and electrodialysis require an external constraint in the form of heat or electrical voltage for transmembrane mass transfer to occur, forward osmosis requires only an osmotic pressure gradient. This can be created by choosing a suitable draw solution that allows water to pass through the membrane, while solutes such as electrolytes are retained by the selective-acting membrane. Thus, forward osmosis alone does not require external energy to achieve permeate flux.
However, for a continuous process commonly used in industry, a steady supply of a highly concentrated draw solution is required. This can come from a continuous secondary process or, after forward osmosis where it has been diluted by the permeate, it must be brought to its initial concentration in a further process step. For the aforementioned reconcentration, any process can be considered. Often these are pressure-driven processes such as reverse osmosis, nanofiltration or ultrafiltration, but combinations with membrane distillation are also possible. Compared to direct concentration of the feed without forward osmosis, the hybrid variant offers the advantage that the feed solution is chosen by the user and is therefore free of impurities. This results in longer lifetimes and more cost-efficient operating points that can be achieved.
MionTec GmbH, as a specialized service provider for ion exchange resins, piloting and laboratory services, has in the past increasingly researched processes with forward osmosis and has already been able to sell its first systems to industrial customers. The company continuously collects measurement data from the plants and can remotely control the plant if necessary. This allows for experimentation with different modes of operation, providing valuable data for forward osmosis optimization. In the near future, the company would like to further expand the field of pressureless membrane processes, and forward osmosis in particular, so that it becomes self-sustaining without the funded projects that have been necessary to date. This requires a fast, cost-efficient and systematized handling of inquiries from potential customers.
MionTec GmbH has already built up an empirical database through years of experience in the field of ion exchange resins, which supports it in the assessment of inquiries through simulations and the cost-optimized control of its own ion exchange resin plants through supervised learning algorithms. The same is now to be achieved for forward osmosis hybrid processes.
Initial problem
The prerequisite for the use of forward osmosis is that it should reduce the operating costs of the overall process in such a way that the additional investment costs pay for themselves over the lifetime of the plant. In the simplest case, forward osmosis is added before an existing/established process such as reverse osmosis. It is then up to the user to decide on a draw solution. If a process is newly designed, any treatment process (ultrafiltration, nanofiltration, membrane distillation, etc.) can also be chosen for the draw solution. Based on the many degrees of freedom available, there are a number of decisions that the process engineer must make in advance for the design, such as:
- Regarding forward osmosis
- Membrane material
- Membrane geometry
- Membrane orientation
- Membrane area
- Module size
- Module geometry
- ...
- Regarding the treatment process (membrane process)
- Process type
- Process variant
- Procedure modification
- Membrane material
- Membrane geometry
- Membrane orientation
- Membrane area
- Module size
- Module geometry
- ...
- Regarding the draw solution
- Type of substance(s) dissolved
- Number of substances dissolved
- Mixture of the dissolved substances
- Concentration of the dissolved substances
- ...
- Regarding the process conditions
- Volume flow of feed solution
- Volume flow of draw solution
- Temperature of feed solution
- Temperature of draw solution
- Pressure (if pressure driven preparation process)
- ...
- Regarding rinsing and cleaning conditions
- Type of flushing and cleaning
- Frequency of flushing and cleaning
- Duration of flushing and cleaning
- Selection of the cleaning agent
- ...
- Regarding the design of the plant
- ...
The number of possible combinations for the process along with the lack of application knowledge regarding forward osmosis is a major problem for many engineers. Not only is it so difficult to build a pilot-scale plant that will perform as desired on a sustained basis, but it is especially problematic to persuade customers to purchase such a plant unless it is proven that the technology will not perform at least as economically as competing processes.
This requires a number of quantitative parameters that should be used to assess techno-economic feasibility. For example, fouling parameters (e.g. MFI-UF, ORI, etc.), which reflect membrane wear, reverse solute flux (RSF), which represents the loss of solutes from the draw solution, or specific energy demand, which represents the amount of energy required to produce one cubic meter of permeate, are discussed for this purpose. These characteristic values can be measured on a laboratory scale with reasonable effort. At pilot plant scale, however, such large quantities of raw solution are required that usually only 1-2 tests can be carried out with the customer's sample. Larger-scale tests such as a statistical design of experiments or long-term studies are impossible and are then postponed to the piloting phase, which in turn is linked to high investment costs.
Research Object
The aim of the research project is to shorten the development time of forward osmosis hybrid processes from the customer inquiry to a pilot plant. The aim is to clarify whether pilot plant tests can be dispensed with entirely when designing the plant if, in addition to geometric parameters (e.g. membrane geometry), the process parameters (e.g. volume flow rates), chemical compositions and thermodynamic properties of the feed and draw solution are also known.
The first step is to clarify whether the behavior of mixtures of substances with a known chemical composition in forward osmosis can be modeled with sufficient accuracy. To this end, meaningful characteristics of forward osmosis are to be simulated and subsequently measured in the pilot plant.
Another important point that requires a lot of development time is the selection of a suitable draw solution. This must not only meet the requirements of forward osmosis, but must also be easily reconcentrated by the downstream treatment step. Based on permeate performance and unwanted reverse diffusion through the forward osmosis membrane, it is not possible to directly identify which draw solution is the most economical. Due to their molecular size, organic salts offer the advantage of having a low RSF and thus require less frequent replenishment. Therefore, they will now be investigated in more detail as draw solvents.