(Bio)electrochemical recovery of ammonia from urine

Abstract

Nitrogen removal from wastewaters is necessary to prevent the pollution of receiving water bodies. At the same time, nitrogen is an essential nutrient for plants, so it is used in the production of fertilizers. Both the conventional removal of nitrogen compounds from wastewater and their production are energy intensive. For this reason, recovering nitrogen directly from wastewater, instead of removing it, can result in reduced energy consumption associated with both its production and removal processes. The use of electrochemical systems (ES) and bioelectrochemical systems (BES) for the recovery of ammonia from wastewaters has been investigated over the past few years. These systems have been proposed as a sustainable alternative to conventional processes because they have the potential to recover energy (in the form of electricity or H2) from wastewaters while recovering ammonia. From all the domestic wastewater streams, urine contains most of the nitrogen (around 80%), and represents only 1% of the volume. Urine can be collected by the use of urine-diverting toilets or waterless urinals, preventing the dilution of the nutrients with high volumes of potable water. In this thesis, we evaluated the use of (bio)electrochemical systems for the recovery of total ammonia nitrogen (TAN) from urine. Our focus was on improving the understanding and performance of the system in terms of TAN recovery. This was needed to get the technology closer to application as, at the start of this research, the proof-of-principle had just been demonstrated, and the highest TAN recoveries were around 30%. The feasibility of BES as a technology for energy-efficient TAN recovery was evaluated in Chapter 2. It was shown that BESs can become economically feasible if, on top of electricity or hydrogen production, the benefits of TAN removal are taken into account. According to our analysis, when TAN removal is taken into account, BESs can still be economic at high internal resistances (200 mΩ.m2), which makes it easier for the application of the system at a bigger scale. The need to develop and test a method to effectively extract ammonia from the catholyte solution was identified as one of the main limiting factors of the system. One of the main conclusions from Chapter 2 was that it was crucial to couple an effective TAN stripping system to the BES to increase TAN recovery. This was addressed in Chapter 3, and followed up in Chapters 4 and 5. Finally, it was determined that to improve TAN recovery in BES, the interactions between the factors affecting the recovery of TAN in BES (such as current density, catholyte pH, TAN concentration, etc.) should be studied in more detail. In Chapter 3, we demonstrated that an ES with an integrated gas-permeable hydrophobic membrane unit can effectively recover TAN from urine. Furthermore, the relationship between current density and TAN loading rate was studied in more detail. It was shown that the relationship between the applied current density and the TAN loading, here called the load ratio, is essential to assess the TAN removal efficiency and energy input of (B)ESs. The load ratio is useful to find the conditions in which a (B)ES for the recovery of TAN can be operated optimally, but it does not take into account other parameters essential to assess the performance of (B)ESs. These limitations are discussed throughout the thesis and in Chapter 6. In Chapter 4, a hydrogen-recycling electrochemical system (HRES) was developed to minimize the energy input of electrochemical systems for the recovery of TAN. In this technology, the hydrogen gas produced at the cathode is reused as the electron donor in the anode, allowing for ammonia recovery at high rates and relatively low energy input. This technology can be applied to recover TAN from wastewaters that do not contain enough organic matter to be treated in a BES. Furthermore, it lowers the risk of chloride oxidation, which typically occurs in electrochemical systems treating wastewaters with high concentrations of chloride, such as urine. Chloride oxidation can result in the formation of harmful compounds such as chlorine gas, chlorination byproducts and adsorbable organic halides (AOX). At a load ratio of 1.2-1.3, the system accomplished TAN removal efficiencies of 73-82% and recoveries of 60-73%. Additional hydrogen needed to be supplied by a supporting electrolyzer, which accounted for 4-9% of the total energy demand of the system. Finally, in Chapter 5, we tested the concept of load ratio in a BES. The load ratio can be manipulated either by changing the current density or the TAN loading rate. In a BES the current density cannot be controlled as easily as in an electrochemical system, because it depends on the oxidation of organic matter by microorganisms. At the same time, manipulating the TAN loading rate would directly affect the organic loading rate, and therefore the current density. We ran the BES coupled to a gas-permeable hydrophobic membrane on synthetic urine and urine. Both influents were fed at different dilutions, flow rates and certain modifications (such as removing TAN from urine prior to feeding it to the cell) in order to obtain a variety of load ratios. We found out that there was a clear increasing trend in TAN removal efficiency with respect to load ratio for both human and synthetic urine. We did not obtain load ratios higher than 0.8, which means that the current was not enough to transport all the TAN across the membrane. This resulted in overall low removal efficiencies (2 to 47%, with 3 exceptions from 52- 61%). In Chapter 6, we discuss what may cause these low current densities. In this last chapter, we also focus on the reasons why (B)ESs have not been applied at larger scales yet, and give future perspectives and recommendations to bring this promising technology closer to application.