Feasibility study on addressing the global water and energy challenges
15 January 2019
Figure 1: The difference between FO, PRO and RO, in terms of solvent flow, hydraulic pressure applied, and membrane orientation. Source: (Achilli, et al., 2009)
Kevin O’Donovan, a chemical and biopharmaceutical engineering student at Cork Institute of Technology (CIT), describes the final-year research project he undertook, in which the feasibility of the use of aquaporin membranes for osmotically driven processes, specifically Pressure Retarded Osmosis (PRO) and Forward Osmosis (FO), was investigated.
Osmotically driven processes have been identified as possible solutions to certain challenges associated with the generation of both electricity and drinking water. The osmotically driven processes which were specifically focused on in this project are Pressure Retarded Osmosis (PRO) and Forward Osmosis (FO).
PRO is a process which harnesses salinity gradient energy (the energy that exists between water bodies of different salt concentration). In PRO, a water body of low salt concentration and low pressure (feed solution) flows into a module where a semipermeable membrane separates it from a pressurised water body of higher salt concentration (draw solution).
Pure water is drawn across the membrane from the feed solution to the draw solution due to the higher salt concentration of the draw solution. This increases the volumetric flowrate in the draw side of the module, and this increase in volumetric flowrate can be used to generate electricity using a turbine.
FO is a similar process to PRO, except that the draw solution is not pressurised. FO possesses the potential to be an energy efficient and environmentally friendly method of desalinating seawater, potentially desalinating water using far less energy than the most common existing methods.
There are currently no full-scale PRO power plants or FO seawater desalination plants in operation. This is mainly due to the lack of commercially available membranes with a sufficient water permeability and salt rejection characteristics to make the processes economically viable.
One possible solution to the problems with membrane performance is through the use of Aquaporin based FO and PRO membranes. Aquaporin based membranes have been developed in recent years, which can potentially achieve much higher water fluxes and salt rejection than conventional membranes.
Recently a commercially available flat sheet ‘Aquaporin Inside’ membrane has been produced by a Denmark based company called Aquaporin.
The purpose of this project was to investigate the performance of this Aquaporin membrane, and determine if the performance could possibly relate to economically viable full scale FO and PRO plants in the future.
The goals of the project were as follows:
1.) Investigate the performance of the flat sheet ‘Aquaporin Inside’ membrane for FO, using an experimental FO rig based in CIT.
2.) Compare the performance of the ‘Aquaporin Inside’ membrane with an alternative FO membrane.
3.) Experimentally determine the important characteristics of the membrane using a novel technique which requires FO experimentation only.
4.) Mathematically model the performance of the membrane for both FO and PRO applications, using the determined membrane characteristics.
Characterisation of the membrane and mathematical modelling
It was originally planned to conduct both FO and PRO experimentation using the CIT lab rig, but due to the limited amount of membranes available, and the fact that the hydraulic pressure involved in PRO could damage the membrane, it was decided to only conduct FO experimentation.
This meant that an accurate mathematical model would have to be developed in order to predict the performance of the membrane for PRO applications.
An integral part of developing a reliable mathematical model is accurately determining the characteristics of the membrane.
A novel method was used to characterise the membrane, which involved FO experimentation only.
This method claims to be more accurate than the conventional methods, which include a Reverse Osmosis step to determine membrane characteristics.
The determined membrane characteristics are presented in Table 1.
The accuracy of the determined characteristics was analysed by conducting FO experiments for various feed solution concentration and draw solution concentration configurations, and comparing the experimental results with a mathematical model developed using the membrane characteristics presented in Table 1.
Figure 3 presents a comparison between the mathematical model and the experimental results at a feed concentration of 0.05M and 0.1M, and a draw concentration ranging from 0.5M to 3M.
The membrane characteristics determined were quite accurate, with a correlation of 0.99 for figure 3(a) and 0.98 for figure 3(b).
The high accuracy of the determined membrane characteristics meant that a reliable mathematical model of the performance of the membrane for PRO could be developed.
The performance of the membrane for two possible practical applications of full scale PRO plant was modelled.
One application uses seawater as the draw solution, and the other application is uses seawater reverse osmosis (SWRO) brine as the draw solution.
The mathematical model determined the water flux (L/m2.h) and the power density (W/m2) that would be achievable for each application, over a range of hydraulic pressures applied to the draw solution.
A benchmark of 5W/m2 is generally set for PRO membrane power densities, as this is the value at which PRO can be considered economically viable (Achilli, et al., 2009).
It can be seen that SWRO brine achieves a much higher power density than seawater, when used as the draw solution.
Figure 4 shows that the performance of the Aquaporin membrane in terms of power density is far below the benchmark of 5W/m2, with a maximum power density of 0.91W/m2 achievable when using SWRO brine as the draw solution, and 0.23W/m2 when using seawater as the draw solution.
This leads to the conclusion that the ‘Aquaporin Inside’ flat sheet membrane is not a suitable candidate for the membrane to be used in a full-scale PRO plant.
Comparison between performance of ‘Aquaporin Inside’ flat sheet membrane and an ‘Ederna’ spiral wound FO membrane
A comparison was made between the performance of the ‘Aquaporin Inside’ membrane and an ‘Ederna’ FO membrane located in Teagasc, Moorepark. The comparison was made for a practical application of FO, the concentration of milk. For the comparison, a feed solution of Simulated Milk Ultrafiltrate (SMUF) was used, and a draw solution of 2M NaCl was used.
Figure 5 shows that the Ederna membrane has a superior performance to the Aquaporin Inside membrane.
This indicates that the Aquaporin Inside flat sheet membrane is also not a viable membrane option for FO to develop to a full-scale process.
Conclusions and recommendations
• Based on the mathematical model developed for the performance of the ‘Aquaporin Inside’ membrane, it can be concluded that the membrane is not a viable option for use as a largescale PRO membrane, due to the low power dnsities achievable;
• The ‘Aquaporin Inside’ flat sheet membrane has a performance which makes it infeasible for use in a FO desalination process. This conclusion is backed up by the fact that the Ederna membrane had a superior performance when tested under the same operating conditions as the Aquaporin membrane;
• The membrane characterisation method utilised can be considered an accurate method of determining membrane characteristics;
• Future work in this area should be conducted using a membrane which is suitable for PRO and can achieve a much higher water flux than the ‘Aquaporin Inside’ flat sheet membrane used in this study.
I would like to acknowledge the following people for their help throughout the project.
• Dr Aisling O’ Gorman, my academic supervisor, for taking the time to assist and guide me throughout the duration of the project.
• Dr Phil Kelly, for his guidance throughout the project, and for organising the visit to Moorepark, and the use of their forward osmosis rig.
1.) Achilli, A., Cath, T. & Childress, A., 2009. Power generation with pressure retarded osmois: An experimental and theoretical investigation. Journal of Membrane Science, pp. 42-52.