In today’s industry ammonia is produced on a large scale, and is used in the production of fertilisers, cleaning products and much more. Eventually the ammonia enters our wastewater, either directly or through urine. This is then removed with conventional nitrifying-denitrifying treatment or with Anammox. That process consumes energy and does not recover the ammonia.
Therefore, the goal of the ‘From Pollutant to Power’ project, of which this thesis is a part, is to achieve a paradigm shift for ammonia to become an energy resource instead of a wastewater pollutant. The objective of the project is to recover ammonia and use it in a solid oxide fuel cell (SOFC) to produce energy. To do this, a vapour with at least a mass percentage of ammonia (m%NH3) of 5% is needed. The energy produced with the SOFC can then be used to cover the energy need for the extraction of ammonia. The SOFC can produce 4.2MJ·kg-N-1 of thermal energy and 8.4 MJ·kg-N-1 of electrical energy.
This study examines the use of vacuum membrane stripping (VMS) to strip ammonia from wastewater. The research objective is to go from using VMS to recover ammonia from synthetic wastewater to strip ammonia from real wastewater, in order to produce ammonia fuel for a SOFC. To accomplish this, some knowledge gaps need to be filled. First, methods need to be found that concentrate the permeate more than previously possible with the VMS to reach the desired m%NH3 in the permeate. Therefore, the effect of the cross-flow of the feed side was investigated and how to concentrate the permeate vapour with the help of condensation. Next, the effect of contaminants on the VMS process was examined by looking at the effect of salts in the wastewater and testing with real industrial ammonia-rich wastewater from amide production, which contains some organics. The final issue to consider was whether this process would be feasible in terms of energy; therefore, an energy balance was made.
Multiple experimental tests were conducted with VMS to investigate the knowledge gaps mentioned above. The selected test conditions were a minimum pH of 10 to exclude CO2 gas transfer, and to ensure the total ammonia nitrogen (TAN) in the water was in the form of ammonia and not ammonium. For the experiments, feed solutions of ammonium hydroxide and ammonium bicarbonate were selected with a concentration of 1.5, 12 and 20 gTAN·L-1 with a temperature of 35°C. For the cross-flow experiments, the initial test was carried out with the Reynolds numbers 200, 300, 400, 500 and 600. From this initial test it was concluded that a Reynolds number of 200 is laminar and 500 is turbulent, and those values were used in the rest of the experiments. For the experiments with condensation, a condensation tube with cooling water of 5-10°C, 15-20°C and 25-30°C was added to the setup. Next, the composition of the industrial wastewater was investigated. In the end, the results from the different experiments were used to make an energy balance.
The results of the experiments with VMS show that turbulent flow is preferred above laminar flow, due to the increased ammonia flux and the ammonia selectivity of the membrane, which both provide a higher concentration of ammonia and more permeate. Salt (high ionic strength) will lower the ammonia flux through the membrane and the selectivity of the membrane. The condensation experiment was successful and adding a subsequent condensation step will concentrate the ammonia in the permeate vapour. Moreover, this condensation step will make it possible to recover some of the energy lost in the VMS. The experiment with the real wastewater was however not successful. The membrane fouled badly due to the organics in the water, where organics acted as surfactants and decreased the hydrophobicity of the membranes. The energy balance shows that the system would be energetically beneficial in terms of electrical energy with a concentration above 12 gTAN·L-1 in the solution for both ammonium hydroxide and ammonium bicarbonate, and that a concentration around 100 gTAN·L-1 from ammonium hydroxide is needed to make it energy efficient in terms of thermal energy.
More investigation is needed to be able to use real wastewater in a VMS module. For the wastewater tested in this study, stripping with VMS is not an alternative. This does not mean that other wastewater cannot be used in the VMS, therefore, other industrial wastewaters, reject water and urine should be investigated. Another possibility could be to review other stripping methods for wastewaters with relatively high contents of organics.